@article{
	11580_106932,
	author = {Palmieri, Jozsef and Di Lillo, Paolo and Lippi, Martina and Chiaverini, Stefano and Marino, Alessandro},
	title = {A Control Architecture for Safe Trajectory Generation in Human–Robot Collaborative Settings},
	year = {2024},
	journal = {IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING},
	abstract = {This paper introduces a control architecture that enables a robotic system to ensure the safety of human operators entering its workspace. The proposed method utilizes an appropriate metric to measure safety levels and adjusts the robot’s motion to maintain this metric above a minimum threshold. To guarantee safety, the robot scales down and deviates from its intended path. For redundant robots, internal motion is exploited to enhance safety levels further. The approach is incorporated into a Hierarchical Quadratic Programming control framework, allowing the robot to address other control objectives simultaneously, such as handling joint limits. Experimental results with a dual-arm mobile robot developed as part of the EU-funded CANOPIES project demonstrate the effectiveness of the proposed method. Note to Practitioners —This paper was motivated by the problem of ensuring human safety in unstructured environments shared with human operators. We propose a control architecture that allows complex dual-arm robotic systems to operate effectively in such scenarios. The devised architecture gives the robot the capability to slow down a trajectory to follow as well as to deviate from a nominal path to keep a human operator safe. We tested the devised approach in a precision farming setting; however, it can be adopted in any human-robot interaction scenario.},
	keywords = {Human–robot interaction, human safety, dual-arm system, trajectory scaling, path modification, precision agriculture},
	url = {https://ieeexplore.ieee.org/document/10413980},
	doi = {10.1109/TASE.2024.3350976}
}

This paper introduces a control architecture that enables a robotic system to ensure the safety of human operators entering its workspace. The proposed method utilizes an appropriate metric to measure safety levels and adjusts the robot’s motion to maintain this metric above a minimum threshold. To guarantee safety, the robot scales down and deviates from its intended path. For redundant robots, internal motion is exploited to enhance safety levels further. The approach is incorporated into a Hierarchical Quadratic Programming control framework, allowing the robot to address other control objectives simultaneously, such as handling joint limits. Experimental results with a dual-arm mobile robot developed as part of the EU-funded CANOPIES project demonstrate the effectiveness of the proposed method. Note to Practitioners —This paper was motivated by the problem of ensuring human safety in unstructured environments shared with human operators. We propose a control architecture that allows complex dual-arm robotic systems to operate effectively in such scenarios. The devised architecture gives the robot the capability to slow down a trajectory to follow as well as to deviate from a nominal path to keep a human operator safe. We tested the devised approach in a precision farming setting; however, it can be adopted in any human-robot interaction scenario.

@article{
	11580_108423,
	author = {Di Libero, Tommaso and Carissimo, Chiara and Cerro, Gianni and Abbatecola, Angela Marie and Marino, Alessandro and Miele, Gianfranco and Ferrigno, Luigi and Rodio, Angelo},
	title = {An Overall Automated Architecture Based on the Tapping Test Measurement Protocol: Hand Dexterity Assessment through an Innovative Objective Method},
	year = {2024},
	journal = {SENSORS},
	volume = {24},
	abstract = {: The present work focuses on the tapping test, which is a method that is commonly used in the literature to assess dexterity, speed, and motor coordination by repeatedly moving fingers, performing a tapping action on a flat surface. During the test, the activation of specific brain regions enhances fine motor abilities, improving motor control. The research also explores neuromuscular and biomechanical factors related to finger dexterity, revealing neuroplastic adaptation to repetitive movements. To give an objective evaluation of all cited physiological aspects, this work proposes a measurement architecture consisting of the following: (i) a novel measurement protocol to assess the coordinative and conditional capabilities of a population of participants; (ii) a suitable measurement platform, consisting of synchronized and non-invasive inertial sensors to be worn at finger level; (iii) a data analysis processing stage, able to provide the final user (medical doctor or training coach) with a plethora of useful information about the carried-out tests, going far beyond state-of-the-art results from classical tapping test examinations. Particularly, the proposed study underscores the importance interdigital autonomy for complex finger motions, despite the challenges posed by anatomical connections; this deepens our understanding of upper limb coordination and the impact of neuroplasticity, holding significance for motor abilities assessment, improvement, and therapeutic strategies to enhance finger precision. The proof-of-concept test is performed by considering a population of college students. The obtained results allow us to consider the proposed architecture to be valuable for many application scenarios, such as the ones related to neurodegenerative disease evolution monitoring.},
	keywords = {IMU sensors; coordinative abilities; hand dexterity; measurement platform; neuroplasticity; tapping test},
	doi = {10.3390/s24134133},
	number = {13}
}

: The present work focuses on the tapping test, which is a method that is commonly used in the literature to assess dexterity, speed, and motor coordination by repeatedly moving fingers, performing a tapping action on a flat surface. During the test, the activation of specific brain regions enhances fine motor abilities, improving motor control. The research also explores neuromuscular and biomechanical factors related to finger dexterity, revealing neuroplastic adaptation to repetitive movements. To give an objective evaluation of all cited physiological aspects, this work proposes a measurement architecture consisting of the following: (i) a novel measurement protocol to assess the coordinative and conditional capabilities of a population of participants; (ii) a suitable measurement platform, consisting of synchronized and non-invasive inertial sensors to be worn at finger level; (iii) a data analysis processing stage, able to provide the final user (medical doctor or training coach) with a plethora of useful information about the carried-out tests, going far beyond state-of-the-art results from classical tapping test examinations. Particularly, the proposed study underscores the importance interdigital autonomy for complex finger motions, despite the challenges posed by anatomical connections; this deepens our understanding of upper limb coordination and the impact of neuroplasticity, holding significance for motor abilities assessment, improvement, and therapeutic strategies to enhance finger precision. The proof-of-concept test is performed by considering a population of college students. The obtained results allow us to consider the proposed architecture to be valuable for many application scenarios, such as the ones related to neurodegenerative disease evolution monitoring.

@conference{
	11580_101803,
	author = {Di Libero, T. and Carissimo, C. and Cerro, G. and Abbatecola, A. M. and Marino, A. and Miele, G. and Ferrigno, L. and Rodio, A.},
	title = {Motor abilities analysis using a standardized tapping test enhanced by a detailed processing stage: gender and age comparison},
	year = {2023},
	booktitle = {2023 IEEE International Symposium on Medical Measurements and Applications, MeMeA 2023},
	abstract = {Wearable sensors have now become an essential reality in sports to assess motor gesture characteristics and health care in diagnosing movement-related diseases. Inertial devices can be used to reliably test an individual’s motor abilities, avoiding human operator-related errors and providing an automated, accurate and reproducible assessment methodology. This paper proposes a new data processing method for assessing upper limb dexterity abilities based on the standard tapping test protocol using a single IMU sensor. The tapping test is a neuromotor examination that measures a subject’s speed and handling of motor abilities. The purpose of this study was to collect normative data for a sensor-assisted tapping test to evaluate its performance against a standardized method. The study was conducted on 53 students aged 21-30 years. Tapping motion was measured with an IMU sensor placed on the left and right index fingers during five 10-second trials. Differences were found between the genders, dominant and non-dominant hands, number of taps, amplitude range, and intertemps. With the sensor, efficient and accurate assessment of tapping speed and movement kinetics was possible, making it useful for scientific and clinical investigations of motor function. Future purposes include replication of this work on patients with neurodegenerative diseases. This assessment method could help track the course of the disease and the effectiveness of the various treatments patients undergo.},
	keywords = {IMU sensor, wearable device, physical activity, motor abilities, tapping test, coordinative abilities},
	doi = {10.1109/MeMeA57477.2023.10171922},
	isbn = {978-1-6654-9384-0},	
	pages = {1--6}
}

Wearable sensors have now become an essential reality in sports to assess motor gesture characteristics and health care in diagnosing movement-related diseases. Inertial devices can be used to reliably test an individual’s motor abilities, avoiding human operator-related errors and providing an automated, accurate and reproducible assessment methodology. This paper proposes a new data processing method for assessing upper limb dexterity abilities based on the standard tapping test protocol using a single IMU sensor. The tapping test is a neuromotor examination that measures a subject’s speed and handling of motor abilities. The purpose of this study was to collect normative data for a sensor-assisted tapping test to evaluate its performance against a standardized method. The study was conducted on 53 students aged 21-30 years. Tapping motion was measured with an IMU sensor placed on the left and right index fingers during five 10-second trials. Differences were found between the genders, dominant and non-dominant hands, number of taps, amplitude range, and intertemps. With the sensor, efficient and accurate assessment of tapping speed and movement kinetics was possible, making it useful for scientific and clinical investigations of motor function. Future purposes include replication of this work on patients with neurodegenerative diseases. This assessment method could help track the course of the disease and the effectiveness of the various treatments patients undergo.

@conference{
	11580_101804,
	author = {Carissimo, C. and Cerro, G. and Debelle, H. and Packer, E. and Yarnall, A. J. and Rochester, L. and Alcock, Lisa and Ferrigno, L. and Marino, A. and Di Libero, T. and Del Din, S.},
	title = {Enhancing remote monitoring and classification of motor state in Parkinson’s disease using Wearable Technology and Machine Learning},
	year = {2023},
	booktitle = {2023 IEEE International Symposium on Medical Measurements and Applications, MeMeA 2023},
	abstract = {Parkinson's disease (PD) is a neurodegenerative condition where dopaminergic medication, such as levodopa, is typically used to improve motor symptoms, including mobility. Identifying the impact of levodopa on real-world motor state (e.g. ON/OFF/ DYSKINESIA) is important for both clinicians and people with PD. The aim of the present work was to automatically classify medication states using machine learning models. Continuous 7-day data were collected in 26 people with PD using an Inertial Measurement Unit (IMU) placed on the fifth lumbar vertebrae (L5) level. Over the week, each participant was asked to complete a diary by annotating medication states (off-condition and dyskinesias) with a 30-minute resolution. Diary entries were used as reference labels assigned to the processed IMU data. Two different networks were chosen for the classification: the k-Nearest Neighbors algorithm (kNN) to identify ON-OFF-DYSKINESIA classes and Fine Tree (FT) to identify only OFF and ON classes. Preliminary results demonstrate that IMU data paired with machine learning could accurately classify ON-OFF and DYSKINESIA with 84% accuracy and the ON-OFF states were classified with 95% accuracy. These results are encouraging and pave the way to a better understanding of the effect that medication has on motor symptoms in PD during everyday life and may serve as a useful tool for optimizing clinical management of people with PD.},
	keywords = {experimental campaign; Inertial Measurement Units (IMU); Parkinson's disease; remote healthcare},
	doi = {10.1109/MeMeA57477.2023.10171868},
	isbn = {978-1-6654-9384-0},	
	pages = {1--6}
}

Parkinson's disease (PD) is a neurodegenerative condition where dopaminergic medication, such as levodopa, is typically used to improve motor symptoms, including mobility. Identifying the impact of levodopa on real-world motor state (e.g. ON/OFF/ DYSKINESIA) is important for both clinicians and people with PD. The aim of the present work was to automatically classify medication states using machine learning models. Continuous 7-day data were collected in 26 people with PD using an Inertial Measurement Unit (IMU) placed on the fifth lumbar vertebrae (L5) level. Over the week, each participant was asked to complete a diary by annotating medication states (off-condition and dyskinesias) with a 30-minute resolution. Diary entries were used as reference labels assigned to the processed IMU data. Two different networks were chosen for the classification: the k-Nearest Neighbors algorithm (kNN) to identify ON-OFF-DYSKINESIA classes and Fine Tree (FT) to identify only OFF and ON classes. Preliminary results demonstrate that IMU data paired with machine learning could accurately classify ON-OFF and DYSKINESIA with 84% accuracy and the ON-OFF states were classified with 95% accuracy. These results are encouraging and pave the way to a better understanding of the effect that medication has on motor symptoms in PD during everyday life and may serve as a useful tool for optimizing clinical management of people with PD.

@article{
	11580_101323,
	author = {Carissimo, Chiara and Cerro, Gianni and Libero, Tommaso Di and Ferrigno, Luigi and Marino, Alessandro and Rodio, Angelo},
	title = {Objective Evaluation of Coordinative Abilities and Training Effectiveness in Sports Scenarios: An Automated Measurement Protocol},
	year = {2023},
	journal = {IEEE ACCESS},
	volume = {11},
	abstract = {The monitoring of coordinative abilities in sports applications is often carried out by trainers that adopt subjective protocols and evaluations not supported by repeatable and reproducible measurement setups. This often leads to unreliable evaluations that do not allow us to simply quantify the positive or negative effect of some training. In this scenario, the rapid spreading of wearable devices able to capture human movements providing data to the users could be a useful instrument to face the problem of simplifying the development of automated, repeatable, and reproducible measurement procedures easily adoptable by a community of athletes or coaches. Following this path, the paper proposes an automatic measurement protocol for the assessment of coordinative abilities based on the use of IMUs embedded in wearable devices. A new protocol based on ruler and tapping tests and a set of objective key performance indicators, derived from IMU measurements, to evaluate the outcome of the test is then developed. In detail, the protocol is based on the sequence 1) ruler test; 2) tapping test; 3) ruler test. The tapping test is performed until energy exhaustion to try and identify, from inertial data, features that can describe possible fatigue effects and correlations with reaction time. Ruler tests are adopted to evaluate the reaction time. The first ruler test provides reaction time information in rest conditions, while the last one considers it after a repeated movement. The comparison of these two times will show whether the reaction time changes after a fatigue condition. An algorithm capable of calculating the number of tapping and the reaction time of each subject is implemented to evaluate the accelerometric data acquired during the tests. Therefore, the impact of the work is two-fold: from an engineering point of view, the automation and performance evaluation of the proposed algorithm is provided; from a sport-medical perspective, the main finding is a general reduction of the reaction time after the energy-exhausting tapping test, as if this last could be considered as a powerful warm-up exercise for sports people. Two are the main results achieved: i) the proposed protocol allows a reduction of the reactions time in about the 83% of cases; ii) the proposed measurement systems allows obtaining, regarding the tapping test, additional very useful quantities as the tap intertemps, frequency spectra, and acceleration excursions, which typically are not provided in state-of-the-art tapping test execution.},
	keywords = {Tapping Test, IMU Sensors, Reaction Time, Viscoelasticity, Measurement Protocol},
	doi = {10.1109/ACCESS.2023.3290471},
	pages = {76996--77008}
}

The monitoring of coordinative abilities in sports applications is often carried out by trainers that adopt subjective protocols and evaluations not supported by repeatable and reproducible measurement setups. This often leads to unreliable evaluations that do not allow us to simply quantify the positive or negative effect of some training. In this scenario, the rapid spreading of wearable devices able to capture human movements providing data to the users could be a useful instrument to face the problem of simplifying the development of automated, repeatable, and reproducible measurement procedures easily adoptable by a community of athletes or coaches. Following this path, the paper proposes an automatic measurement protocol for the assessment of coordinative abilities based on the use of IMUs embedded in wearable devices. A new protocol based on ruler and tapping tests and a set of objective key performance indicators, derived from IMU measurements, to evaluate the outcome of the test is then developed. In detail, the protocol is based on the sequence 1) ruler test; 2) tapping test; 3) ruler test. The tapping test is performed until energy exhaustion to try and identify, from inertial data, features that can describe possible fatigue effects and correlations with reaction time. Ruler tests are adopted to evaluate the reaction time. The first ruler test provides reaction time information in rest conditions, while the last one considers it after a repeated movement. The comparison of these two times will show whether the reaction time changes after a fatigue condition. An algorithm capable of calculating the number of tapping and the reaction time of each subject is implemented to evaluate the accelerometric data acquired during the tests. Therefore, the impact of the work is two-fold: from an engineering point of view, the automation and performance evaluation of the proposed algorithm is provided; from a sport-medical perspective, the main finding is a general reduction of the reaction time after the energy-exhausting tapping test, as if this last could be considered as a powerful warm-up exercise for sports people. Two are the main results achieved: i) the proposed protocol allows a reduction of the reactions time in about the 83% of cases; ii) the proposed measurement systems allows obtaining, regarding the tapping test, additional very useful quantities as the tap intertemps, frequency spectra, and acceleration excursions, which typically are not provided in state-of-the-art tapping test execution.

@conference{
	11580_104484,
	author = {Golluccio, Giacomo and Di Lillo, Paolo and Marino, Alessandro and Antonelli, Gianluca},
	title = {When Local Optimization is Bad: Learning What to (Not) Maximize in the Null-Space for Redundant Robot Control},
	year = {2023},
	publisher = {IEEE},
	booktitle = {IEEE International conference on control, decision and information technologies},
	keywords = {robotics},
	doi = {10.1109/CoDIT58514.2023.10284456},
	isbn = {979-8-3503-1140-2},	
	pages = {506--511}
}

@conference{
	11580_106724,
	author = {Lippi, Martina and Di Lillo, Paolo and Marino, Alessandro},
	title = {A Task Allocation Framework for Human Multi-Robot Collaborative Settings},
	year = {2023},
	publisher = {IEEE},
	volume = {2023-May},
	booktitle = {IEEE INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION},
	abstract = {The requirements of modern production systems together with more advanced robotic technologies have fostered the integration of teams comprising humans and autonomous robots. While this integration has the potential to provide various benefits, it also raises questions about how to effectively manage these teams, taking into account the different characteristics of the agents involved. This paper presents a framework for task allocation in a human multi-robot collaborative scenario. The proposed solution combines an optimal offline allocation with an online reallocation strategy which accounts for inaccuracies of the offline plan and/or unforeseen events, human subjective preferences and cost of task switching. Experiments with two manipulators cooperating with a human operator in a box filling task are presented.},
	keywords = {Production systems, Human-Robot collabroation, Precision Agriculture},
	url = {https://ieeexplore.ieee.org/abstract/document/10161458},
	doi = {10.1109/icra48891.2023.10161458},
	isbn = {979-8-3503-2366-5},	
	pages = {7614--7620}
}

The requirements of modern production systems together with more advanced robotic technologies have fostered the integration of teams comprising humans and autonomous robots. While this integration has the potential to provide various benefits, it also raises questions about how to effectively manage these teams, taking into account the different characteristics of the agents involved. This paper presents a framework for task allocation in a human multi-robot collaborative scenario. The proposed solution combines an optimal offline allocation with an online reallocation strategy which accounts for inaccuracies of the offline plan and/or unforeseen events, human subjective preferences and cost of task switching. Experiments with two manipulators cooperating with a human operator in a box filling task are presented.

@conference{
	11580_106723,
	author = {Lippi, Martina and Gallou, Jorand and Gasparri, Andrea and Marino, Alessandro},
	title = {An Optimal Allocation and Scheduling Method in Human-Multi-Robot Precision Agriculture Settings},
	year = {2023},
	publisher = {IEEE},
	booktitle = {31st Mediterranean Conference on Control and Automation, MED 2022},
	abstract = {Employing teams of robots to offer services to human operators enables the latter to reduce their physical workload. In this paper, we focus on the problem of optimally allocating and scheduling the robot tasks in order to serve human operators. We formulate a Mixed-Integer Linear Programming problem that aims to minimize the human waiting time and the energy spent by the robots, while ensuring that any velocity constraints of the robots are fulfilled and the task ordering is correct. In addition, we propose an online re-allocation strategy that takes into account the possibility of changing human parameters over time. This strategy determines whether a new optimal solution must be computed. We validate the proposed framework in a simulated precision agriculture setting composed of two robots and four human operators for a harvesting application.},
	keywords = {Human-robot collaboration, Optimal Allocation, Precision Agriculture},
	url = {https://ieeexplore.ieee.org/abstract/document/10185899/authors},
	doi = {10.1109/med59994.2023.10185899},
	isbn = {979-8-3503-1544-8},	
	pages = {541--546}
}

Employing teams of robots to offer services to human operators enables the latter to reduce their physical workload. In this paper, we focus on the problem of optimally allocating and scheduling the robot tasks in order to serve human operators. We formulate a Mixed-Integer Linear Programming problem that aims to minimize the human waiting time and the energy spent by the robots, while ensuring that any velocity constraints of the robots are fulfilled and the task ordering is correct. In addition, we propose an online re-allocation strategy that takes into account the possibility of changing human parameters over time. This strategy determines whether a new optimal solution must be computed. We validate the proposed framework in a simulated precision agriculture setting composed of two robots and four human operators for a harvesting application.

@conference{
	11580_109743,
	author = {Furchì, Antonio and Lippi, Martina and Marino, Alessandro and Gasparri, Andrea},
	title = {Distributed Finite-Time Supremum/Infimum Dynamic Consensus Under Directed Network Topology},
	year = {2023},
	booktitle = {IEEE Conference on Decision and Control (CDC)},
	abstract = {In this paper, we address the distributed supremum/infimum dynamic consensus problem in networked multi-agent systems. More in detail, by considering that each agent has access to a local exogenous time-varying signal, the objective is to have all the agents distributively track the global maximum supremum (or minimum infimum) of these exogenous signals. We propose a distributed protocol guaranteeing finite-time convergence under directed network topology. The sole requirements are the strong connectivity of the communication graph and the boundedness of the derivatives of the exogenous signals, with known bounds. The effectiveness of the proposed protocol is corroborated through numerical simulations in a precision farming case study.},
	keywords = {Distributed control; multi-agent systems.},
	url = {https://ieeexplore.ieee.org/document/10384085/authors#authors},
	doi = {10.1109/cdc49753.2023.10384085},
	isbn = {979-8-3503-0124-3},	
	pages = {4480--4485}
}

In this paper, we address the distributed supremum/infimum dynamic consensus problem in networked multi-agent systems. More in detail, by considering that each agent has access to a local exogenous time-varying signal, the objective is to have all the agents distributively track the global maximum supremum (or minimum infimum) of these exogenous signals. We propose a distributed protocol guaranteeing finite-time convergence under directed network topology. The sole requirements are the strong connectivity of the communication graph and the boundedness of the derivatives of the exogenous signals, with known bounds. The effectiveness of the proposed protocol is corroborated through numerical simulations in a precision farming case study.

@conference{
	11580_109744,
	author = {Lippi, M. and Gallou, J. and Palmieri, J. and Gasparri, A. and Marino, A.},
	title = {Human-Multi-Robot Task Allocation in Agricultural Settings: a Mixed Integer Linear Programming Approach},
	year = {2023},
	publisher = {IEEE},
	booktitle = {IEEE International Conference on Robot and Human Interactive Communication (RO-MAN)},
	abstract = {The use of heterogeneous human-multi-robot teams enables the combination of complementary skills of these two different types of agents. To have an effective collaboration, it is necessary to define a strategy for allocating and scheduling tasks among them. In this work, we distinguish robots in working robots and service ones: working robots and human operators can perform similar tasks in the environment and both are assisted by service robots. We propose a Mixed-Integer Linear Programming approach that aims to minimize the waiting times of the working agents, the energy consumption of the service robots, and the makespan while ensuring that the velocity constraints of the robots are met and the task ordering is correct. Furthermore, we propose an online updating strategy that tackles changes in the parameters of working agents and adapts the plan accordingly based on a heuristic algorithm. To validate our framework, we analyze a precision agriculture harvesting application with two human operators, two working robots, and two service robots.},
	keywords = {Human-Robot collaboration; task allocation},
	url = {https://ieeexplore.ieee.org/document/10309392},
	doi = {10.1109/RO-MAN57019.2023.10309392},
	isbn = {979-8-3503-3670-2},	
	pages = {1056--1062}
}

The use of heterogeneous human-multi-robot teams enables the combination of complementary skills of these two different types of agents. To have an effective collaboration, it is necessary to define a strategy for allocating and scheduling tasks among them. In this work, we distinguish robots in working robots and service ones: working robots and human operators can perform similar tasks in the environment and both are assisted by service robots. We propose a Mixed-Integer Linear Programming approach that aims to minimize the waiting times of the working agents, the energy consumption of the service robots, and the makespan while ensuring that the velocity constraints of the robots are met and the task ordering is correct. Furthermore, we propose an online updating strategy that tackles changes in the parameters of working agents and adapts the plan accordingly based on a heuristic algorithm. To validate our framework, we analyze a precision agriculture harvesting application with two human operators, two working robots, and two service robots.

@article{
	11580_93946,
	author = {Lippi, M. and Poklukar, P. and Welle, M. C. and Varava, A. and Yin, H. and Marino, A. and Kragic, D.},
	title = {Enabling Visual Action Planning for Object Manipulation Through Latent Space Roadmap},
	year = {2022},
	journal = {IEEE TRANSACTIONS ON ROBOTICS},
	abstract = {In this article, we present a framework for visual action planning of complex manipulation tasks with high-dimensional state spaces, focusing on manipulation of deformable objects. We propose a latent space roadmap (LSR) for task planning, which is a graph-based structure globally capturing the system dynamics in a low-dimensional latent space. Our framework consists of the following three parts. First, a mapping module (MM) that maps observations is given in the form of images into a structured latent space extracting the respective states as well as generates observations from the latent states. Second, the LSR, which builds and connects clusters containing similar states in order to find the latent plans between start and goal states, extracted by MM. Third, the action proposal module that complements the latent plan found by the LSR with the corresponding actions. We present a thorough investigation of our framework on simulated box stacking and rope/box manipulation tasks, and a folding task executed on a real robot.},
	keywords = {Deep Learning in Robotics and Automation; Heuristic algorithms; Latent Space Planning; Manipulation Planning; Planning; Robots; Stacking; Task analysis; Trajectory; Visual Learning; Visualization},
	url = {https://ieeexplore.ieee.org/document/9833914},
	doi = {10.1109/TRO.2022.3188163},
	pages = {1--19},
	number = {1}
}

In this article, we present a framework for visual action planning of complex manipulation tasks with high-dimensional state spaces, focusing on manipulation of deformable objects. We propose a latent space roadmap (LSR) for task planning, which is a graph-based structure globally capturing the system dynamics in a low-dimensional latent space. Our framework consists of the following three parts. First, a mapping module (MM) that maps observations is given in the form of images into a structured latent space extracting the respective states as well as generates observations from the latent states. Second, the LSR, which builds and connects clusters containing similar states in order to find the latent plans between start and goal states, extracted by MM. Third, the action proposal module that complements the latent plan found by the LSR with the corresponding actions. We present a thorough investigation of our framework on simulated box stacking and rope/box manipulation tasks, and a folding task executed on a real robot.

@article{
	11580_92238,
	author = {Lippi, Martina and Furchi, Antonio and Marino, Alessandro and Gasparri, Andrea},
	title = {An Adaptive Distributed Protocol for Finite-time Infimum or Supremum Dynamic Consensus},
	year = {2022},
	journal = {IEEE CONTROL SYSTEMS LETTERS},
	abstract = {In this paper, the problem of distributively tracking the minimum infimum (or maximum supremum) of a set of time-varying signals in finite-time is addressed. More specifically, each agent has access to a local time-varying exogenous signal, and all the agents are required to follow the minimum infimum (or the maximum supremum) of these signals in a distributed fashion. No assumption is made on the network size nor on the bounds of the exogenous signal derivatives. An adaptive protocol is provided which can provably solve the above problem in finite-time for multi-agent systems with undirected connected network topologies. Numerical simulations are provided to corroborate the theoretical findings.},
	keywords = {Distributed Control, Adaptive Control, Dynamic Consensus, Multi-Agent Systems},
	url = {https://ieeexplore.ieee.org/document/9816009},
	doi = {10.1109/LCSYS.2022.3188941},
	pages = {1--6}
}

In this paper, the problem of distributively tracking the minimum infimum (or maximum supremum) of a set of time-varying signals in finite-time is addressed. More specifically, each agent has access to a local time-varying exogenous signal, and all the agents are required to follow the minimum infimum (or the maximum supremum) of these signals in a distributed fashion. No assumption is made on the network size nor on the bounds of the exogenous signal derivatives. An adaptive protocol is provided which can provably solve the above problem in finite-time for multi-agent systems with undirected connected network topologies. Numerical simulations are provided to corroborate the theoretical findings.

@conference{
	11580_98686,
	author = {Carissimo, C. and Ferrigno, L. and Golluccio, G. and Marino, A. and Cerro, G.},
	title = {Parkinson's disease aided diagnosis: online symptoms detection by a low-cost wearable Inertial Measurement Unit},
	year = {2022},
	publisher = {Institute of Electrical and Electronics Engineers Inc.},
	booktitle = {2022 IEEE International Symposium on Medical Measurements and Applications, MeMeA 2022 - Conference Proceedings},
	abstract = {The usage of mini-devices in medicine for continuous non-invasive monitoring of neurodegenerative diseases is rapidly increasing. Among most common diseases belonging to such category, Parkinson's is one of the main disorders, especially in aged population. It is characterized by several symptoms whose comprehensive and accurate analysis can lead to a punctual and effective diagnosis. This task is generally accomplished by an expert medical doctor but, especially in first stage, the aid of an automatic tool can help to catch even very low symptomatology. A promising solution to detect most motor issues related to Parkinson's disease is represented by Inertial Measurement Units (IMUs), typically including accelerometers, magnetometers and gyroscopes. Their metrological features, such as accuracy, sensitivity and immunity to external disturbances are critical to get a fully functional and discriminant device. Furthermore, the capability to extrapolate pathological states from measurements is a very attractive feature to automatize early warning and fast medical interventions. To accomplish for both tasks, in this paper a measuring platform containing an IMU is presented and metrologically characterized; moreover, classification tests for typical impairments due to Parkinson's disease are proposed. Although improvements in the procedure and measurement quality are on the way, the current status allows to state its suitability for the required application framework.},
	keywords = {IMU sensor; machine learning; Parkinson's disease; tremor measurement; wearable},
	doi = {10.1109/MeMeA54994.2022.9856546},
	isbn = {978-1-6654-8299-8},	
	pages = {1--6}
}

The usage of mini-devices in medicine for continuous non-invasive monitoring of neurodegenerative diseases is rapidly increasing. Among most common diseases belonging to such category, Parkinson's is one of the main disorders, especially in aged population. It is characterized by several symptoms whose comprehensive and accurate analysis can lead to a punctual and effective diagnosis. This task is generally accomplished by an expert medical doctor but, especially in first stage, the aid of an automatic tool can help to catch even very low symptomatology. A promising solution to detect most motor issues related to Parkinson's disease is represented by Inertial Measurement Units (IMUs), typically including accelerometers, magnetometers and gyroscopes. Their metrological features, such as accuracy, sensitivity and immunity to external disturbances are critical to get a fully functional and discriminant device. Furthermore, the capability to extrapolate pathological states from measurements is a very attractive feature to automatize early warning and fast medical interventions. To accomplish for both tasks, in this paper a measuring platform containing an IMU is presented and metrologically characterized; moreover, classification tests for typical impairments due to Parkinson's disease are proposed. Although improvements in the procedure and measurement quality are on the way, the current status allows to state its suitability for the required application framework.

@article{
	11580_98684,
	author = {Carissimo, Chiara and Cerro, Gianni and Ferrigno, Luigi and Golluccio, Giacomo and Marino, Alessandro},
	title = {Development and Assessment of a Movement Disorder Simulator Based on Inertial Data},
	year = {2022},
	journal = {SENSORS},
	volume = {22},
	abstract = {: The detection analysis of neurodegenerative diseases by means of low-cost sensors and suitable classification algorithms is a key part of the widely spreading telemedicine techniques. The choice of suitable sensors and the tuning of analysis algorithms require a large amount of data, which could be derived from a large experimental measurement campaign involving voluntary patients. This process requires a prior approval phase for the processing and the use of sensitive data in order to respect patient privacy and ethical aspects. To obtain clearance from an ethics committee, it is necessary to submit a protocol describing tests and wait for approval, which can take place after a typical period of six months. An alternative consists of structuring, implementing, validating, and adopting a software simulator at most for the initial stage of the research. To this end, the paper proposes the development, validation, and usage of a software simulator able to generate movement disorders-related data, for both healthy and pathological conditions, based on raw inertial measurement data, and give tri-axial acceleration and angular velocity as output. To present a possible operating scenario of the developed software, this work focuses on a specific case study, i.e., the Parkinson's disease-related tremor, one of the main disorders of the homonym pathology. The full framework is reported, from raw data availability to pathological data generation, along with a common machine learning method implementation to evaluate data suitability to be distinguished and classified. Due to the development of a flexible and easy-to-use simulator, the paper also analyses and discusses the data quality, described with typical measurement features, as a metric to allow accurate classification under a low-performance sensing device. The simulator's validation results show a correlation coefficient greater than 0.94 for angular velocity and 0.93 regarding acceleration data. Classification performance on Parkinson's disease tremor was greater than 98% in the best test conditions.},
	keywords = {IMU data; Parkinson’s disease; machine learning; measurement; simulation; tremor detection},
	doi = {10.3390/s22176341},
	number = {17}
}

: The detection analysis of neurodegenerative diseases by means of low-cost sensors and suitable classification algorithms is a key part of the widely spreading telemedicine techniques. The choice of suitable sensors and the tuning of analysis algorithms require a large amount of data, which could be derived from a large experimental measurement campaign involving voluntary patients. This process requires a prior approval phase for the processing and the use of sensitive data in order to respect patient privacy and ethical aspects. To obtain clearance from an ethics committee, it is necessary to submit a protocol describing tests and wait for approval, which can take place after a typical period of six months. An alternative consists of structuring, implementing, validating, and adopting a software simulator at most for the initial stage of the research. To this end, the paper proposes the development, validation, and usage of a software simulator able to generate movement disorders-related data, for both healthy and pathological conditions, based on raw inertial measurement data, and give tri-axial acceleration and angular velocity as output. To present a possible operating scenario of the developed software, this work focuses on a specific case study, i.e., the Parkinson's disease-related tremor, one of the main disorders of the homonym pathology. The full framework is reported, from raw data availability to pathological data generation, along with a common machine learning method implementation to evaluate data suitability to be distinguished and classified. Due to the development of a flexible and easy-to-use simulator, the paper also analyses and discusses the data quality, described with typical measurement features, as a metric to allow accurate classification under a low-performance sensing device. The simulator's validation results show a correlation coefficient greater than 0.94 for angular velocity and 0.93 regarding acceleration data. Classification performance on Parkinson's disease tremor was greater than 98% in the best test conditions.

@conference{
	11580_98683,
	author = {Lippi, M. and Welle, M. C. and Poklukar, P. and Marino, A. and Kragic, D.},
	title = {Augment-Connect-Explore: a Paradigm for Visual Action Planning with Data Scarcity},
	year = {2022},
	publisher = {Institute of Electrical and Electronics Engineers Inc.},
	volume = {2022-},
	booktitle = {IEEE International Conference on Intelligent Robots and Systems},
	abstract = {Visual action planning particularly excels in applications where the state of the system cannot be computed explicitly, such as manipulation of deformable objects, as it enables planning directly from raw images. Even though the field has been significantly accelerated by deep learning techniques, a crucial requirement for their success is the availability of a large amount of data. In this work, we propose the Augment-Connect-Explore (ACE) paradigm to enable visual action planning in cases of data scarcity. We build upon the Latent Space Roadmap (LSR) framework which performs planning with a graph built in a low dimensional latent space. In particular, ACE is used to i) Augment the available training dataset by autonomously creating new pairs of datapoints, ii) create new unobserved Connections among representations of states in the latent graph, and iii) Explore new regions of the latent space in a targeted manner. We validate the proposed approach on both simulated box stacking and real-world folding task showing the applicability for rigid and deformable object manipulation tasks, respectively.},
	url = {https://ieeexplore.ieee.org/abstract/document/9982199/authors#authors},
	doi = {10.1109/IROS47612.2022.9982199},
	isbn = {978-1-6654-7927-1},	
	pages = {754--761}
}

Visual action planning particularly excels in applications where the state of the system cannot be computed explicitly, such as manipulation of deformable objects, as it enables planning directly from raw images. Even though the field has been significantly accelerated by deep learning techniques, a crucial requirement for their success is the availability of a large amount of data. In this work, we propose the Augment-Connect-Explore (ACE) paradigm to enable visual action planning in cases of data scarcity. We build upon the Latent Space Roadmap (LSR) framework which performs planning with a graph built in a low dimensional latent space. In particular, ACE is used to i) Augment the available training dataset by autonomously creating new pairs of datapoints, ii) create new unobserved Connections among representations of states in the latent graph, and iii) Explore new regions of the latent space in a targeted manner. We validate the proposed approach on both simulated box stacking and real-world folding task showing the applicability for rigid and deformable object manipulation tasks, respectively.

@conference{
	11580_98685,
	author = {Lippi, Martina and Furchi, Antonio and Marino, Alessandro and Gasparri, Andrea},
	title = {Finite-Time Distributed Protocol for Tracking the Upper (Lower) Bound For a Set of Time-Varying Reference Signals},
	year = {2022},
	booktitle = {30th Mediterranean Conference on Control and Automation, MED 2022},
	abstract = {In this paper we address the problem of distributively tracking the upper or lower bound of n time-varying signals in finite-time. In detail, each agent has access to a time-varying exogenous signal, which may encode the evolution of a physical phenomenon. All the agents are required to follow the upper (lower) bound value among such output signals in a distributed fashion. We provide a protocol to solve the above problem in the case of networked agents and undirected communication, together with formal proof of convergence and estimation of an upperbound of the convergence time. We corroborate the protocol via numerical validations in a precision farming setting.},
	doi = {10.1109/MED54222.2022.9837239},
	isbn = {978-1-6654-0673-4},	
	pages = {908--913}
}

In this paper we address the problem of distributively tracking the upper or lower bound of n time-varying signals in finite-time. In detail, each agent has access to a time-varying exogenous signal, which may encode the evolution of a physical phenomenon. All the agents are required to follow the upper (lower) bound value among such output signals in a distributed fashion. We provide a protocol to solve the above problem in the case of networked agents and undirected communication, together with formal proof of convergence and estimation of an upperbound of the convergence time. We corroborate the protocol via numerical validations in a precision farming setting.

@article{
	11580_93947,
	author = {Golluccio, G. and Di Lillo, P. and Di Vito, D. and Marino, A. and Antonelli, G.},
	title = {Objects Relocation in Clutter with Robot Manipulators via Tree-based Q-Learning Algorithm: Analysis and Experiments},
	year = {2022},
	journal = {JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS},
	volume = {106},
	abstract = {This work addresses the problem of retrieving a target object from cluttered environment using a robot manipulator. In the details, the proposed solution relies on a Task and Motion Planning approach based on a two-level architecture: the high-level is a Task Planner aimed at finding the optimal objects sequence to relocate, according to a metric based on the objects weight; the low-level is a Motion Planner in charge of planning the end-effector path for reaching the specific objects taking into account the robot physical constraints. The high-level task planner is a Reinforcement Learning agent, trained using the information coming from the low-level Motion Planner. In this work we consider the Q-Tree algorithm, which is based on a dynamic tree structure inspired by the Q-learning technique. Three different RL-policies with two kinds of tree exploration techniques (Breadth and Depth) are compared in simulation scenarios with different complexity. Moreover, the proposed learning methods are experimentally validated in a real scenario by adopting a KINOVA Jaco2 7-DoFs robot manipulator.},
	keywords = {Motion planning; Reinforcement learning; Task planning},
	url = {https://link.springer.com/article/10.1007/s10846-022-01719-9},
	doi = {10.1007/s10846-022-01719-9},
	number = {2}
}

This work addresses the problem of retrieving a target object from cluttered environment using a robot manipulator. In the details, the proposed solution relies on a Task and Motion Planning approach based on a two-level architecture: the high-level is a Task Planner aimed at finding the optimal objects sequence to relocate, according to a metric based on the objects weight; the low-level is a Motion Planner in charge of planning the end-effector path for reaching the specific objects taking into account the robot physical constraints. The high-level task planner is a Reinforcement Learning agent, trained using the information coming from the low-level Motion Planner. In this work we consider the Q-Tree algorithm, which is based on a dynamic tree structure inspired by the Q-learning technique. Three different RL-policies with two kinds of tree exploration techniques (Breadth and Depth) are compared in simulation scenarios with different complexity. Moreover, the proposed learning methods are experimentally validated in a real scenario by adopting a KINOVA Jaco2 7-DoFs robot manipulator.

@article{
	11580_79469,
	author = {Gillini, Giuseppe and Di Lillo, Paolo and Arrichiello, Filippo and Di Vito, Daniele and Marino, Alessandro and Antonelli, Gianluca and Chiaverini, Stefano},
	title = {A dual-arm mobile robot system performing assistive tasks operated via P300-based brain computer interface},
	year = {2022},
	journal = {INDUSTRIAL ROBOT},
	volume = {49},
	keywords = {Control, Robotics, Man–machine interface (MMI), Autonomous robots},
	doi = {10.1108/IR-07-2020-0137},
	pages = {11--20},
	number = {1}
}

@conference{
	11580_92618,
	author = {Lippi, Martina and Marino, Alessandro},
	title = {A mixed-integer linear programming formulation for human multi-robot task allocation},
	year = {2021},
	publisher = {IEEE},
	booktitle = {30th IEEE International Conference on Robot & Human Interactive Communication (RO-MAN)},
	abstract = {In this work, we address a task allocation problem for human multi-robot settings. Given a set of tasks to perform, we formulate a general Mixed-Integer Linear Programming (MILP) problem aiming at minimizing the overall execution time while optimizing the quality of the executed tasks as well as human and robotic workload. Different skills of the agents, both human and robotic, are taken into account and human operators are enabled to either directly execute tasks or play supervisory roles; moreover, multiple manipulators can tightly collaborate if required to carry out a task. Finally, as realistic in human contexts, human parameters are assumed to vary over time, e.g., due to increasing human level of fatigue. Therefore, online monitoring is required and re-allocation is performed if needed. Simulations in a realistic scenario with two manipulators and a human operator performing an assembly task validate the effectiveness of the approach.},
	doi = {10.1109/RO-MAN50785.2021.9515362},
	isbn = {978-1-6654-0492-1},	
	pages = {1017--1023}
}

In this work, we address a task allocation problem for human multi-robot settings. Given a set of tasks to perform, we formulate a general Mixed-Integer Linear Programming (MILP) problem aiming at minimizing the overall execution time while optimizing the quality of the executed tasks as well as human and robotic workload. Different skills of the agents, both human and robotic, are taken into account and human operators are enabled to either directly execute tasks or play supervisory roles; moreover, multiple manipulators can tightly collaborate if required to carry out a task. Finally, as realistic in human contexts, human parameters are assumed to vary over time, e.g., due to increasing human level of fatigue. Therefore, online monitoring is required and re-allocation is performed if needed. Simulations in a realistic scenario with two manipulators and a human operator performing an assembly task validate the effectiveness of the approach.

@conference{
	11580_92258,
	author = {Golluccio, Giacomo and DI VITO, Daniele and Marino, Alessandro and Antonelli, Gianluca},
	title = {Robotic Weight-based Object Relocation in Clutter via Tree-based Q-learning Approach using Breadth and Depth Search Techniques},
	year = {2021},
	publisher = {IEEE},
	booktitle = {2021 20th International Conference on Advanced Robotics (ICAR)},
	abstract = {In this paper, the problem of retrieving a target object from a cluttered environment through a mobile manipulator is considered. The task is solved by combining Task and Motion Planning; in detail, at a higher level, the task planner is in charge of planning the sequence of objects to relocate while, at a lower level, the motion planner is in charge of planning the robot movements taking into consideration robot and environment constraints. In particular, the latter provides feedback to the former about the feasibility of object sequences; this information is exploited to train a Reinforcement Learning agent that, according to an objects-weight based metrics, builds a dynamic decision-tree where each node represents a sequence of relocated objects, and edge values are weights updated via Q-learning-inspired algorithm. Three learning strategies differing in how the tree is explored are analysed. Moreover, each exploration approach is performed using two different tree-search methods: the Breadth first and Depth first techniques. Finally, the proposed learning strategies are numerically validated and compared in three scenarios of growing-complexity.},
	keywords = {Measurement
Q-learning
Heuristic algorithms
Search problems
Manipulators
Learning (Artificial Intelligence)
Mobile robots
Path planning
Tree searching
Planning
Trajectory},
	url = {https://ieeexplore.ieee.org/abstract/document/9659471},
	doi = {10.1109/ICAR53236.2021.9659471},
	isbn = {978-1-6654-3684-7},	
	pages = {676--681}
}

In this paper, the problem of retrieving a target object from a cluttered environment through a mobile manipulator is considered. The task is solved by combining Task and Motion Planning; in detail, at a higher level, the task planner is in charge of planning the sequence of objects to relocate while, at a lower level, the motion planner is in charge of planning the robot movements taking into consideration robot and environment constraints. In particular, the latter provides feedback to the former about the feasibility of object sequences; this information is exploited to train a Reinforcement Learning agent that, according to an objects-weight based metrics, builds a dynamic decision-tree where each node represents a sequence of relocated objects, and edge values are weights updated via Q-learning-inspired algorithm. Three learning strategies differing in how the tree is explored are analysed. Moreover, each exploration approach is performed using two different tree-search methods: the Breadth first and Depth first techniques. Finally, the proposed learning strategies are numerically validated and compared in three scenarios of growing-complexity.

@conference{
	11580_89183,
	author = {Lippi, M. and Marino, A.},
	title = {A control barrier function approach to human-multi-robot safe interaction},
	year = {2021},
	publisher = {Institute of Electrical and Electronics Engineers Inc.},
	booktitle = {29th Mediterranean Conference on Control and Automation, MED 2021},
	abstract = {A fundamental prerequisite to achieve a successful human-robot cooperation is human safety, which becomes even more crucial when multiple robots are involved in the cooperative task. A general solution for addressing safety in human-multi-robot scenarios is proposed in this paper. Human safety is assessed by a safety field which accounts for the multi-robot system as a whole. The assessment of human safety is exploited within an optimization protocol where the robots' cooperative task trajectory is scaled whenever required, while exploiting the system redundancy. Control Barrier Functions (CBFs) are adopted to set up the optimization problem and kinematic and dynamic constraints are taken into consideration. Finally, simulations on a realistic setup composed of three industrial mobile manipulators show the effectiveness of the proposed solution.},
	url = {https://ieeexplore.ieee.org/abstract/document/9480187},
	doi = {10.1109/MED51440.2021.9480187},
	isbn = {978-1-6654-2258-1},	
	pages = {604--609}
}

A fundamental prerequisite to achieve a successful human-robot cooperation is human safety, which becomes even more crucial when multiple robots are involved in the cooperative task. A general solution for addressing safety in human-multi-robot scenarios is proposed in this paper. Human safety is assessed by a safety field which accounts for the multi-robot system as a whole. The assessment of human safety is exploited within an optimization protocol where the robots' cooperative task trajectory is scaled whenever required, while exploiting the system redundancy. Control Barrier Functions (CBFs) are adopted to set up the optimization problem and kinematic and dynamic constraints are taken into consideration. Finally, simulations on a realistic setup composed of three industrial mobile manipulators show the effectiveness of the proposed solution.

@article{
	11580_77405,
	author = {Lippi, Martina and Marino, Alessandro},
	title = {Human Multi-Robot Safe Interaction: A Trajectory Scaling Approach Based on Safety Assessment},
	year = {2021},
	journal = {IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY},
	abstract = {Nowadays, industrial robotics requires that robots and humans share the same workspace and collaborate to a certain extent. In such a scenario, the safety is the minimum requirement and, for this reason, many off-the-shelf collaborative robots are now available on the market which, basically, are able to limit the contact forces in the case of impact. Differently from many different works, this article presents a solution to human multi-robot safe interaction in which multiple mobile manipulators are in charge of performing a cooperative task in a workspace shared with human operators. The safety of the interaction is assessed by a safety field that considers the whole system and is general enough concerning its expression. Based on the value of this field, the cooperative task trajectory is properly modified so as to ensure a safe interaction while trying to preserve as much as possible the nominal task, which is instead completely aborted whenever the safety of the interaction cannot be guaranteed. The solution is first designed within a centralized architecture and, then, upon this, a distributed implementation is presented, which, in general, aims to exhibit the same performance as the centralized counterpart. Finally, both simulations and experiments on real industrial robots corroborate the designed solution.},
	url = {https://ieeexplore.ieee.org/document/9146918},
	doi = {10.1109/TCST.2020.3009031},
	pages = {1--16},
	number = {4}
}

Nowadays, industrial robotics requires that robots and humans share the same workspace and collaborate to a certain extent. In such a scenario, the safety is the minimum requirement and, for this reason, many off-the-shelf collaborative robots are now available on the market which, basically, are able to limit the contact forces in the case of impact. Differently from many different works, this article presents a solution to human multi-robot safe interaction in which multiple mobile manipulators are in charge of performing a cooperative task in a workspace shared with human operators. The safety of the interaction is assessed by a safety field that considers the whole system and is general enough concerning its expression. Based on the value of this field, the cooperative task trajectory is properly modified so as to ensure a safe interaction while trying to preserve as much as possible the nominal task, which is instead completely aborted whenever the safety of the interaction cannot be guaranteed. The solution is first designed within a centralized architecture and, then, upon this, a distributed implementation is presented, which, in general, aims to exhibit the same performance as the centralized counterpart. Finally, both simulations and experiments on real industrial robots corroborate the designed solution.

@article{
	11580_80455,
	author = {Lippi, Martina and Marino, Alessandro},
	title = {Human Multi-Robot Physical Interaction: a Distributed Framework},
	year = {2021},
	journal = {JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS},
	volume = {101},
	abstract = {The objective of this paper is to devise a general framework to allow a human operator to physically interact with an object
manipulated by a multi-manipulator system in a distributed setting. A two layer solution is devised. In detail, at the top layer
an arbitrary virtual dynamics is considered for the object with the virtual input chosen as the solution of an optimal Linear
Quadratic Tracking (LQT) problem. In this formulation, both the human and robots’ intentions are taken into account, being
the former online estimated by Recursive Least Squares (RLS) technique. The output of this layer is a desired trajectory of
the object which is the input of the bottom layer and from which desired trajectories for the robot end effectors are computed
based on the closed-chain constraints. Each robot, then, implements a time-varying gain adaptive control law so as to take
into account model uncertainty and internal wrenches that inevitably raise due to synchronization errors and dynamic and
kinematic uncertainties. Remarkably, the overall solution is devised in a distributed setting by resorting to a leader-follower
approach and distributed observers. Simulations with three 6-DOFs serial chain manipulators mounted on mobile platforms
corroborate the theoretical findings.},
	keywords = {Human-robot interaction; Shared control; Distributed control},
	url = {https://link.springer.com/article/10.1007/s10846-020-01277-y},
	doi = {10.1007/s10846-020-01277-y},
	pages = {1--20},
	number = {2}
}

The objective of this paper is to devise a general framework to allow a human operator to physically interact with an object manipulated by a multi-manipulator system in a distributed setting. A two layer solution is devised. In detail, at the top layer an arbitrary virtual dynamics is considered for the object with the virtual input chosen as the solution of an optimal Linear Quadratic Tracking (LQT) problem. In this formulation, both the human and robots’ intentions are taken into account, being the former online estimated by Recursive Least Squares (RLS) technique. The output of this layer is a desired trajectory of the object which is the input of the bottom layer and from which desired trajectories for the robot end effectors are computed based on the closed-chain constraints. Each robot, then, implements a time-varying gain adaptive control law so as to take into account model uncertainty and internal wrenches that inevitably raise due to synchronization errors and dynamic and kinematic uncertainties. Remarkably, the overall solution is devised in a distributed setting by resorting to a leader-follower approach and distributed observers. Simulations with three 6-DOFs serial chain manipulators mounted on mobile platforms corroborate the theoretical findings.

@conference{
	11580_89185,
	author = {Lippi, M. and Gillini, G. and Marino, A. and Arrichiello, F.},
	title = {A Data-Driven Approach for Contact Detection, Classification and Reaction in Physical Human-Robot Collaboration},
	year = {2021},
	publisher = {Institute of Electrical and Electronics Engineers Inc.},
	booktitle = {Proceedings - IEEE International Conference on Robotics and Automation},
	abstract = {This paper considers a scenario where a robot and a human operator share the same workspace, and the robot is able to both carry out autonomous tasks and physically interact with the human in order to achieve common goals. In this context, both intentional and accidental contacts between human and robot might occur due to the complexity of tasks and environment, to the uncertainty of human behavior, and to the typical lack of awareness of each other actions. Here, a two stage strategy based on Recurrent Neural Networks (RNNs) is designed to detect intentional and accidental contacts: the occurrence of a contact with the human is detected at the first stage, while the classification between intentional and accidental is performed at the second stage. An admittance control strategy or an evasive action is then performed by the robot, respectively. The approach also works in the case the robot simultaneously interacts with the human and the environment, where the interaction wrench of the latter is modeled via Gaussian Mixture Models (GMMs). Control Barrier Functions (CBFs) are included, at the control level, to guarantee the satisfaction of robot and task constraints while performing the proper interaction strategy. The approach has been validated on a real setup composed of a Kinova Jaco2 robot.},
	url = {https://ieeexplore.ieee.org/abstract/document/9480187},
	doi = {10.1109/ICRA48506.2021.9561827},
	isbn = {978-1-7281-9077-8},	
	pages = {12730--12736}
}

This paper considers a scenario where a robot and a human operator share the same workspace, and the robot is able to both carry out autonomous tasks and physically interact with the human in order to achieve common goals. In this context, both intentional and accidental contacts between human and robot might occur due to the complexity of tasks and environment, to the uncertainty of human behavior, and to the typical lack of awareness of each other actions. Here, a two stage strategy based on Recurrent Neural Networks (RNNs) is designed to detect intentional and accidental contacts: the occurrence of a contact with the human is detected at the first stage, while the classification between intentional and accidental is performed at the second stage. An admittance control strategy or an evasive action is then performed by the robot, respectively. The approach also works in the case the robot simultaneously interacts with the human and the environment, where the interaction wrench of the latter is modeled via Gaussian Mixture Models (GMMs). Control Barrier Functions (CBFs) are included, at the control level, to guarantee the satisfaction of robot and task constraints while performing the proper interaction strategy. The approach has been validated on a real setup composed of a Kinova Jaco2 robot.

@conference{
	11580_87263,
	author = {Golluccio, G. and Di Vito, D. and Marino, A. and Bria, A. and Antonelli, G.},
	title = {Task-motion planning via tree-based Q-learning approach for robotic object displacement in cluttered spaces},
	year = {2021},
	publisher = {SciTePress},
	booktitle = {Proceedings of the 18th International Conference on Informatics in Control, Automation and Robotics, ICINCO 2021},
	abstract = {In this paper, a Reinforcement Learning approach to the problem of grasping a target object from clutter by a robotic arm is addressed. A layered architecture is devised to the scope. The bottom layer is in charge of planning robot motion in order to relocate objects while taking into account robot constraints, whereas the top layer takes decision about which obstacles to relocate. In order to generate an optimal sequence of obstacles according to some metrics, a tree is dynamically built where nodes represent sequences of relocated objects and edge weights are updated according to a Q-learning-inspired algorithm. Four different exploration strategies of the solution tree are considered, ranging from a random strategy to a ε-Greedy learning-based exploration. The four strategies are compared based on some predefined metrics and in scenarios with different complexity. The learning-based approaches are able to provide optimal relocation sequences despite the high dimensional search space, with the ε-Greedy strategy showing better performance, especially in complex scenarios.},
	keywords = {Motion planning; Reinforcement learning; Task planning},
	doi = {10.5220/0010542601300137},
	isbn = {978-989-758-522-7},	
	pages = {130--137}
}

In this paper, a Reinforcement Learning approach to the problem of grasping a target object from clutter by a robotic arm is addressed. A layered architecture is devised to the scope. The bottom layer is in charge of planning robot motion in order to relocate objects while taking into account robot constraints, whereas the top layer takes decision about which obstacles to relocate. In order to generate an optimal sequence of obstacles according to some metrics, a tree is dynamically built where nodes represent sequences of relocated objects and edge weights are updated according to a Q-learning-inspired algorithm. Four different exploration strategies of the solution tree are considered, ranging from a random strategy to a ε-Greedy learning-based exploration. The four strategies are compared based on some predefined metrics and in scenarios with different complexity. The learning-based approaches are able to provide optimal relocation sequences despite the high dimensional search space, with the ε-Greedy strategy showing better performance, especially in complex scenarios.

@conference{
	11580_91484,
	author = {Lippi, M. and Marino, A.},
	title = {Enabling physical human-robot collaboration through contact classification and reaction},
	year = {2020},
	publisher = {IEEE},
	booktitle = {29th IEEE International Conference on Robot and Human Interactive Communication (RO-MAN)},
	abstract = {In this paper, a scenario of physical human-robot
collaboration is considered, in which a robot is able to both
carry out autonomous tasks and to physically interact with a
human operator to achieve a common objective. However, since
human and robot share the same workspace both accidental
and intentional contacts between them might arise. Therefore,
a solution based on Recurrent Neural Networks (RNNs) is
proposed to detect and classify the nature of the contact with
the human, even in the case the robot is interacting with the
environment because of its own task. Then, reaction strategies
are defined depending on the nature of contact: human avoidance with evasive action in the case of accidental interaction,
and admittance control in the case of intentional interaction.
In regard to the latter, Control Barrier Functions (CBFs) are
considered to guarantee the satisfaction of robot constraints,
while endowing the robot with a compatible compliant behavior.
The approach is validated on real data acquired from the
interaction with a Kinova Jaco2.},
	doi = {10.1109/RO-MAN47096.2020.9223580},
	isbn = {978-1-7281-6075-7},	
	pages = {1196--1203}
}

In this paper, a scenario of physical human-robot collaboration is considered, in which a robot is able to both carry out autonomous tasks and to physically interact with a human operator to achieve a common objective. However, since human and robot share the same workspace both accidental and intentional contacts between them might arise. Therefore, a solution based on Recurrent Neural Networks (RNNs) is proposed to detect and classify the nature of the contact with the human, even in the case the robot is interacting with the environment because of its own task. Then, reaction strategies are defined depending on the nature of contact: human avoidance with evasive action in the case of accidental interaction, and admittance control in the case of intentional interaction. In regard to the latter, Control Barrier Functions (CBFs) are considered to guarantee the satisfaction of robot constraints, while endowing the robot with a compatible compliant behavior. The approach is validated on real data acquired from the interaction with a Kinova Jaco2.

@article{
	11580_74457,
	author = {Garcia-Camacho, Irene and Alenya, Guillem and Kragic, Danica and Lippi, Martina and Welle, Michael C. and Yin, Hang and Antonova, Rika and Varava, Anastasiia and Borras, Julia and Torras, Carme and Marino, Alessandro},
	title = {Benchmarking Bimanual Cloth Manipulation},
	year = {2020},
	journal = {IEEE ROBOTICS AND AUTOMATION LETTERS},
	volume = {5},
	abstract = {Cloth manipulation is a challenging task that, despite its importance, has received relatively little attention compared to rigid object manipulation. In this paper, we provide three benchmarks for evaluation and comparison of different approaches towards three basic tasks in cloth manipulation: spreading a tablecloth over a table, folding a towel, and dressing. The tasks can be executed on any bimanual robotic platform and the objects involved in the tasks are standardized and easy to acquire. We provide several complexity levels for each task, and describe the quality measures to evaluate task execution. Furthermore, we provide baseline solutions for all the tasks and evaluate them according to the proposed metrics.},
	keywords = {Performance Evaluation and Benchmarking, Cooperating Robots},
	doi = {10.1109/LRA.2020.2965891},
	pages = {1111--1118},
	number = {2}
}

Cloth manipulation is a challenging task that, despite its importance, has received relatively little attention compared to rigid object manipulation. In this paper, we provide three benchmarks for evaluation and comparison of different approaches towards three basic tasks in cloth manipulation: spreading a tablecloth over a table, folding a towel, and dressing. The tasks can be executed on any bimanual robotic platform and the objects involved in the tasks are standardized and easy to acquire. We provide several complexity levels for each task, and describe the quality measures to evaluate task execution. Furthermore, we provide baseline solutions for all the tasks and evaluate them according to the proposed metrics.

@article{
	11580_78237,
	author = {Golluccio, G. and Gillini, G. and Marino, A. and Antonelli, G.},
	title = {Robot Dynamics Identification: A Reproducible Comparison With Experiments on the KINOVA Jaco2},
	year = {2020},
	journal = {IEEE ROBOTICS AND AUTOMATION MAGAZINE},
	keywords = {Computational modeling; Data models; Linear matrix inequalities; Mathematical model; Robot sensing systems; Solid modeling},
	doi = {10.1109/MRA.2020.3004149},
	pages = {0--0},
	number = {3}
}

@article{
	11580_73835,
	author = {Gasparri, Andrea and Marino, Alessandro},
	title = {A Distributed Framework for k-hop Control Strategies in Large-Scale Networks Based on Local Interactions},
	year = {2020},
	journal = {IEEE TRANSACTIONS ON AUTOMATIC CONTROL},
	volume = {65},
	abstract = {We propose a distributed framework for large scale networks to attain control strategies requiring k-hop interactions. This research is motivated by the observation that in many practical applications and operational domains involving large-scale networks, such as environmental monitoring or traffic load balancing, agents may be required to collect only information concerning other agents located sufficiently close to them, that is agents topologically at most k-hop away. In this set- ting, distributed observers available at the state of art, which typically estimates the full network state, may be inadequate due to scalability issues. Differently, we propose a distributed finite- time observer which allows each agent to estimate the state of its k-hop neighbors by interacting only with the agents belonging to its 1-hop neighborhood. Furthermore, we demonstrate that for feedback control strategies based on k-hop neighborhood information, which are Input-to-State stable, the proposed distributed finite-time observer can be effectively used to design stable large- scale networked control strategies. Numerical results are provided to corroborate the theoretical findings.},
	url = {https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8754717},
	doi = {10.1109/TAC.2019.2926595},
	pages = {1825--1840},
	number = {5}
}

We propose a distributed framework for large scale networks to attain control strategies requiring k-hop interactions. This research is motivated by the observation that in many practical applications and operational domains involving large-scale networks, such as environmental monitoring or traffic load balancing, agents may be required to collect only information concerning other agents located sufficiently close to them, that is agents topologically at most k-hop away. In this set- ting, distributed observers available at the state of art, which typically estimates the full network state, may be inadequate due to scalability issues. Differently, we propose a distributed finite- time observer which allows each agent to estimate the state of its k-hop neighbors by interacting only with the agents belonging to its 1-hop neighborhood. Furthermore, we demonstrate that for feedback control strategies based on k-hop neighborhood information, which are Input-to-State stable, the proposed distributed finite-time observer can be effectively used to design stable large- scale networked control strategies. Numerical results are provided to corroborate the theoretical findings.

@conference{
	11580_91478,
	author = {Lippi, M. and Poklukar, P. and Welle, M. C. and Varava, A. and Yin, H. and Marino, A. and Kragic, D.},
	title = {Latent space roadmap for visual action planning of deformable and rigid object manipulation},
	year = {2020},
	booktitle = {International Conference on Intelligent Robots and Systems},
	abstract = {We present a framework for visual action planning
of complex manipulation tasks with high-dimensional state
spaces such as manipulation of deformable objects. Planning is
performed in a low-dimensional latent state space that embeds
images. We define and implement a Latent Space Roadmap
(LSR) which is a graph-based structure that globally captures
the latent system dynamics. Our framework consists of two
main components: a Visual Foresight Module (VFM) that
generates a visual plan as a sequence of images, and an Action
Proposal Network (APN) that predicts the actions between
them. We show the effectiveness of the method on a simulated
box stacking task as well as a T-shirt folding task performed
with a real robot.},
	url = {https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9340764},
	doi = {10.1109/IROS45743.2020.9340764},
	isbn = {978-1-7281-6212-6},	
	pages = {5619--5626}
}

We present a framework for visual action planning of complex manipulation tasks with high-dimensional state spaces such as manipulation of deformable objects. Planning is performed in a low-dimensional latent state space that embeds images. We define and implement a Latent Space Roadmap (LSR) which is a graph-based structure that globally captures the latent system dynamics. Our framework consists of two main components: a Visual Foresight Module (VFM) that generates a visual plan as a sequence of images, and an Action Proposal Network (APN) that predicts the actions between them. We show the effectiveness of the method on a simulated box stacking task as well as a T-shirt folding task performed with a real robot.

@inbook{
	11580_77421,
	author = {Gillini, Giuseppe and Lippi, Martina and Arrichiello, Filippo and Marino, Alessandro and Pierri, Francesco},
	title = {Fault Diagnosis and Fault-tolerant Control of Robotic and Autonomous Systems},
	year = {2020},
	publisher = {IET Digital Library},
	booktitle = {Distributed fault detection and isolation strategy for a team of cooperative mobile manipulators},
	abstract = {Applications involving multi-robot systems have been increasing day by day, since they allow one to accomplish complex tasks otherwise impossible for a single unit. Common control approaches for these robot systems are based on distributed architecture, where each robot computes its own control input, only based on local information from onboard sensors or received from its neighbor robots. This means that the failure of one or more agents might jeopardize the task execution. For this reason, fault detection and isolation (FDI) strategies become crucial to accomplish the assigned task in the aforementioned case as well. This chapter presents a distributed fault diagnosis architecture aimed at detecting failures in a team of robots working in tight cooperation. The proposed approach relies on a distributed observer-controller scheme, where each robot estimates the overall system state by means of a local observer, and it uses such an estimate to compute the local control input to achieve a specific task. The local observer is also used to define a set of residual vectors aimed at detecting and isolating faults occurring on any robot of the team, even if there is no direct communication. The approach is validated through experiments where a heterogeneous team of three robots perform a cooperative task.},
	keywords = {Fault diagnosis; multi-robot system, cooperative manipulation; distributed control},
	url = {https://digital-library.theiet.org/content/books/10.1049/pbce126e_ch7},
	doi = {10.1049/PBCE126E},
	isbn = {9781785618307},	
	isbn = {9781785618314},	
	pages = {143--166}
}

Applications involving multi-robot systems have been increasing day by day, since they allow one to accomplish complex tasks otherwise impossible for a single unit. Common control approaches for these robot systems are based on distributed architecture, where each robot computes its own control input, only based on local information from onboard sensors or received from its neighbor robots. This means that the failure of one or more agents might jeopardize the task execution. For this reason, fault detection and isolation (FDI) strategies become crucial to accomplish the assigned task in the aforementioned case as well. This chapter presents a distributed fault diagnosis architecture aimed at detecting failures in a team of robots working in tight cooperation. The proposed approach relies on a distributed observer-controller scheme, where each robot estimates the overall system state by means of a local observer, and it uses such an estimate to compute the local control input to achieve a specific task. The local observer is also used to define a set of residual vectors aimed at detecting and isolating faults occurring on any robot of the team, even if there is no direct communication. The approach is validated through experiments where a heterogeneous team of three robots perform a cooperative task.

@conference{
	11580_74039,
	author = {Gillini, Giuseppe and Lippi, Martina and Arrichiello, Filippo and Marino, Alessandro and Pierri, Francesco},
	title = {Distributed Fault Detection and Isolation for Cooperative Mobile Manipulators},
	year = {2019},
	booktitle = {2019 IEEE International Conference on Systems, Man and Cybernetics (SMC)},
	abstract = {The paper presents a Distributed Fault Detection and Isolation strategy for a team of mobile manipulators performing a cooperative mission. The overall system relies on an observer-controller scheme where each robot estimates the global state of the team through a distributed observer; then, the global state estimate is used by each robot to compute the estimated local input so as to achieve a specific global task. The observer-controller scheme also allows to define a set of residual vectors that can be used by the robots to detect and isolate faults affecting any member of the team, even if not in direct communication, and without increasing the computational burden and the information exchange. The approach is validated via numerical simulations with a team of four mobile manipulators performing a transportation mission.},
	keywords = {Observers, Task analysis, Robot sensing systems, Manipulator dynamics, Fault detection},
	doi = {10.1109/SMC.2019.8913953},
	isbn = {978-1-7281-4569-3},	
	pages = {1701--1707}
}

The paper presents a Distributed Fault Detection and Isolation strategy for a team of mobile manipulators performing a cooperative mission. The overall system relies on an observer-controller scheme where each robot estimates the global state of the team through a distributed observer; then, the global state estimate is used by each robot to compute the estimated local input so as to achieve a specific global task. The observer-controller scheme also allows to define a set of residual vectors that can be used by the robots to detect and isolate faults affecting any member of the team, even if not in direct communication, and without increasing the computational burden and the information exchange. The approach is validated via numerical simulations with a team of four mobile manipulators performing a transportation mission.

@conference{
	11580_74345,
	author = {Lippi, Martina and Marino, Alessandro and Chiaverini, Stefano},
	title = {A distributed approach to human multi-robot physical interaction},
	year = {2019},
	publisher = {Institute of Electrical and Electronics Engineers Inc.},
	volume = {2019-October},
	booktitle = {2019 IEEE International Conference on Systems, Man and Cybernetics (SMC)},
	abstract = {In this paper, a distributed scheme to allow a human operator to physically interact with a multi-manipulator system is devised. Manipulators are tightly connected to a rigid object and a human operator interacts with it to perform, for example, a cooperative transportation task. The strategy foresees two layers. The top layer is in charge of assigning a compliant behaviour to the object through an admittance model whose reference trajectory is dynamically adjusted to regulate the human-object interaction force. Moreover, since the parameters of the dynamic model of the human arm end-point are supposed to be time-varying and completely unknowns with unknown bounds, a robust adaptive control is envisaged in this layer. The output of this layer is a desired object trajectory which is tracked by the bottom layer. In detail, the latter resorts to a robust adaptive control strategy to both track the object trajectory and control the internal stresses exerted by the manipulators on the object which unavoidably arise due to dynamic and kinematic uncertainties and synchronization errors. Simulations involving a setup with three dual-arm Movo mobile robots corroborate the theoretical findings. © 2019 IEEE.},
	keywords = {Adaptive control systems; Manipulators; Personnel; Trajectories, Cooperative transportation; Distributed approaches; Human-object interaction; Kinematic uncertainty; Physical interactions; Reference trajectories; Robust-adaptive control; Synchronization error, Human robot interaction},
	url = {https://www.scopus.com/inward/record.uri?eid=2-s2.0-85076726216&doi=10.1109/SMC.2019.8914468&partnerID=40&md5=03016ba95f9d90d0cdad4d49a8e235b5},
	doi = {10.1109/SMC.2019.8914468},
	isbn = {9781728145693},	
	pages = {728--734}
}

In this paper, a distributed scheme to allow a human operator to physically interact with a multi-manipulator system is devised. Manipulators are tightly connected to a rigid object and a human operator interacts with it to perform, for example, a cooperative transportation task. The strategy foresees two layers. The top layer is in charge of assigning a compliant behaviour to the object through an admittance model whose reference trajectory is dynamically adjusted to regulate the human-object interaction force. Moreover, since the parameters of the dynamic model of the human arm end-point are supposed to be time-varying and completely unknowns with unknown bounds, a robust adaptive control is envisaged in this layer. The output of this layer is a desired object trajectory which is tracked by the bottom layer. In detail, the latter resorts to a robust adaptive control strategy to both track the object trajectory and control the internal stresses exerted by the manipulators on the object which unavoidably arise due to dynamic and kinematic uncertainties and synchronization errors. Simulations involving a setup with three dual-arm Movo mobile robots corroborate the theoretical findings. © 2019 IEEE.

@conference{
	11580_71446,
	author = {Lippi, Martina and Marino, Alessandro},
	title = {Cooperative Object Transportation by Multiple Ground and Aerial Vehicles: Modeling and Planning},
	year = {2018},
	booktitle = {2018 IEEE International Conference on Robotics and Automation (ICRA)},
	abstract = {In this paper the modeling and planning problems of a system composed of multiple ground and aerial robots involved in a transportation task are considered. The ground robots rigidly grasp a load, while the aerial vehicles are attached to the object through non-rigid inextensible cables. The idea behind such a heterogeneous multi-robot system is to benefit of the advantages of both types of robots that might be the precision of ground robots, the increased payload of multiple aerial vehicles and their larger workspace. The overall model of the system is derived and its expression and redundancy are exploited by setting a general constrained optimal planning problem. The problem is herein solved by dynamic programming and simulation results validated the proposed scheme.},
	keywords = {Vehicle dynamics,
Manipulators,
Load modeling,
Unmanned aerial vehicles,
Dynamics},
	url = {https://ieeexplore.ieee.org/document/8460778},
	doi = {10.1109/ICRA.2018.8460778},
	isbn = {978-1-5386-3081-5},	
	pages = {1084--1090}
}

In this paper the modeling and planning problems of a system composed of multiple ground and aerial robots involved in a transportation task are considered. The ground robots rigidly grasp a load, while the aerial vehicles are attached to the object through non-rigid inextensible cables. The idea behind such a heterogeneous multi-robot system is to benefit of the advantages of both types of robots that might be the precision of ground robots, the increased payload of multiple aerial vehicles and their larger workspace. The overall model of the system is derived and its expression and redundancy are exploited by setting a general constrained optimal planning problem. The problem is herein solved by dynamic programming and simulation results validated the proposed scheme.

@conference{
	11580_71255,
	author = {Lippi, Martina and Marino, Alessandro},
	title = {Distributed Kinematic Control and Trajectory Scaling for Multi-Manipulator Systems in Presence of Human Operators},
	year = {2018},
	publisher = {IEEE},
	booktitle = {Proceedings of 26th Mediterranean Conference on Control and Automation},
	abstract = {The objective of the paper is to devise a general framework for handling the human safety in a multi-robot work-cell controlled within a decentralized framework. The paper is motivated by the increasing demand coming from new production paradigms for strict cooperation between humans and robots and for flexibility and robustness provided by decentralized control frameworks. The cell foresees several robots with different assigned roles. In particular, it is supposed that there are worker agents, that are in charge of performing the cooperative manipulation task, and watcher robots, that are in charge of supervising the cell with particular attention to the human safety. The latter is guaranteed by properly modifying the workers' task trajectory according to a state transition strategy that tries to preserve the task path as much as possible. The overall solution is tested via simulations in order to show the effectiveness of results.},
	keywords = {decentralised control, distributed control, manipulators, mobile robots, multi-agent systems,
multi-robot systems, robot kinematics,
trajectory control},
	url = {https://ieeexplore.ieee.org/document/8442476},
	doi = {10.1109/MED.2018.8442476},
	isbn = {978-1-5386-7890-9},	
	pages = {377--382}
}

The objective of the paper is to devise a general framework for handling the human safety in a multi-robot work-cell controlled within a decentralized framework. The paper is motivated by the increasing demand coming from new production paradigms for strict cooperation between humans and robots and for flexibility and robustness provided by decentralized control frameworks. The cell foresees several robots with different assigned roles. In particular, it is supposed that there are worker agents, that are in charge of performing the cooperative manipulation task, and watcher robots, that are in charge of supervising the cell with particular attention to the human safety. The latter is guaranteed by properly modifying the workers' task trajectory according to a state transition strategy that tries to preserve the task path as much as possible. The overall solution is tested via simulations in order to show the effectiveness of results.

@article{
	11580_71234,
	author = {Marino, Alessandro and Pierri, Francesco},
	title = {A two stage approach for distributed cooperative manipulation of an unknown object without explicit communication and unknown number of robots},
	year = {2018},
	journal = {ROBOTICS AND AUTONOMOUS SYSTEMS},
	volume = {103},
	abstract = {This paper proposes a distributed algorithm for cooperatively manipulating an object rigidly grasped by a team of mobile manipulators. In order to increase the flexibility of the multi-robot cell and differently from other approaches, it is assumed that the object is completely unknown and there is not information exchange between robots. The devised strategy includes two stages: at the first stage, each robot estimates the object kinematic and dynamic parameters by applying specific contact wrenches, while, in the second stage, the estimated parameters are exploited within a distributed cooperative control framework that can be adopted, for instance, to control the interaction wrench exerted by the environment on the object or to implement a zero-force control algorithm. In addition to the total absence of communication and differently from existing solutions, the proposed technique assumes that each robot has not knowledge of the number of cooperative agents in the team and, remarkably, it is devised in the 3-dimensional space with the aim of handling both the position and the orientation of the object. Finally, the feasibility of the approach is proven via numerical simulations.},
	keywords = {Distributed estimation; Distributed cooperative manipulation.},
	url = {https://www.sciencedirect.com/science/article/pii/S0921889017307807},
	doi = {10.1016/j.robot.2018.02.007},
	pages = {122--133}
}

This paper proposes a distributed algorithm for cooperatively manipulating an object rigidly grasped by a team of mobile manipulators. In order to increase the flexibility of the multi-robot cell and differently from other approaches, it is assumed that the object is completely unknown and there is not information exchange between robots. The devised strategy includes two stages: at the first stage, each robot estimates the object kinematic and dynamic parameters by applying specific contact wrenches, while, in the second stage, the estimated parameters are exploited within a distributed cooperative control framework that can be adopted, for instance, to control the interaction wrench exerted by the environment on the object or to implement a zero-force control algorithm. In addition to the total absence of communication and differently from existing solutions, the proposed technique assumes that each robot has not knowledge of the number of cooperative agents in the team and, remarkably, it is devised in the 3-dimensional space with the aim of handling both the position and the orientation of the object. Finally, the feasibility of the approach is proven via numerical simulations.

@article{
	11580_71239,
	author = {Marino, Alessandro},
	title = {Distributed Adaptive Control of Networked Cooperative Mobile Manipulators},
	year = {2018},
	journal = {IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY},
	volume = {26},
	abstract = {This paper deals with the networked control of loosely or tightly connected cooperative manipulators in charge of achieving a cooperative task that is specified by means of proper task-oriented variables depending on the full state of the system. Since the full state is not known to robots, a two-layer architecture is designed. At the first level, each arm controller runs a distributed observer that estimates the system state. At the second level, this estimation is adopted to compute the local control input as in the case that a central unit is available. In addition, since the dynamic parameters of the arms might not be perfectly known, the local control law is made adaptive in order to counteract this uncertainty. The designed solution is suitable not only for pure motion coordination tasks but can also be exploited in those cases where a closed kinematic chain is generated by multirobot object manipulation. In this situation, the objective is to both move the object and limit the internal stresses on it that, however, cannot be locally computed. To overcome this issue, the wrench exerted by each robot on the object is decomposed in an external component (contributing to the motion of the object) and in an internal component that is locally estimated and, then, regulated. The approach was validated by simulation with 6-DOF serial chain manipulators mounted on a mobile platform and performing cooperative tasks.},
	keywords = {Cooperative robots; distributed control; networked robots.},
	url = {http://ieeexplore.ieee.org/document/7990577/},
	doi = {10.1109/TCST.2017.2720673},
	pages = {1646--1660},
	number = {5}
}

This paper deals with the networked control of loosely or tightly connected cooperative manipulators in charge of achieving a cooperative task that is specified by means of proper task-oriented variables depending on the full state of the system. Since the full state is not known to robots, a two-layer architecture is designed. At the first level, each arm controller runs a distributed observer that estimates the system state. At the second level, this estimation is adopted to compute the local control input as in the case that a central unit is available. In addition, since the dynamic parameters of the arms might not be perfectly known, the local control law is made adaptive in order to counteract this uncertainty. The designed solution is suitable not only for pure motion coordination tasks but can also be exploited in those cases where a closed kinematic chain is generated by multirobot object manipulation. In this situation, the objective is to both move the object and limit the internal stresses on it that, however, cannot be locally computed. To overcome this issue, the wrench exerted by each robot on the object is decomposed in an external component (contributing to the motion of the object) and in an internal component that is locally estimated and, then, regulated. The approach was validated by simulation with 6-DOF serial chain manipulators mounted on a mobile platform and performing cooperative tasks.

@conference{
	11580_71447,
	author = {Santilli, Matteo and Marino, Alessandro and Gasparri, Andrea},
	title = {A Finite-Time Protocol for Distributed Continuous-Time Optimization of Sum of Locally Coupled Strictly Convex Functions},
	year = {2018},
	publisher = {IEEE},
	booktitle = {2018 IEEE Conference on Decision and Control (CDC)},
	abstract = {In this paper we study a distributed optimization problem for continuous time multi-agent systems. In our setting, the global objective for the multi-agent system is to minimize the sum of locally coupled strictly convex cost functions. Notably, this class of optimization objectives can be used to encode several important problems such as distributed estimation. For this problem setting, we propose a distributed signed gradient descent algorithm, which relies on local observers to retrieve 2-hop state information that are required to compute the descent direction. Adaptive gains for the local observer are introduced to render the convergence independent from: i) the structure of the network topology and ii) the local gains of the per-agent signed gradient-descent update law. The finite-time convergence of the local observer and of the proposed signed gradient descent method is demonstrated. Numerical simulations involving a distributed weighted least-square (WLS) estimation problem, with the aim of identifying in the context of an advanced water management system for precision-farming the soil thermal properties in a large-scale hazelnut orchard, have been proposed to corroborate the theoretical findings.},
	url = {https://ieeexplore.ieee.org/abstract/document/8619315},
	doi = {10.1109/CDC.2018.8619315},
	isbn = {978-1-5386-1395-5},	
	pages = {993--998}
}

In this paper we study a distributed optimization problem for continuous time multi-agent systems. In our setting, the global objective for the multi-agent system is to minimize the sum of locally coupled strictly convex cost functions. Notably, this class of optimization objectives can be used to encode several important problems such as distributed estimation. For this problem setting, we propose a distributed signed gradient descent algorithm, which relies on local observers to retrieve 2-hop state information that are required to compute the descent direction. Adaptive gains for the local observer are introduced to render the convergence independent from: i) the structure of the network topology and ii) the local gains of the per-agent signed gradient-descent update law. The finite-time convergence of the local observer and of the proposed signed gradient descent method is demonstrated. Numerical simulations involving a distributed weighted least-square (WLS) estimation problem, with the aim of identifying in the context of an advanced water management system for precision-farming the soil thermal properties in a large-scale hazelnut orchard, have been proposed to corroborate the theoretical findings.

@conference{
	11580_71444,
	author = {Lippi, Martina and Marino, Alessandro},
	title = {Safety in human-multi robot collaborative scenarios: a trajectory scaling approach},
	year = {2018},
	publisher = {Elsevier B.V.},
	journal = {IFAC-PAPERSONLINE},
	volume = {51},
	booktitle = {12th IFAC Symposium on Robot Control SYROCO},
	abstract = {In this paper, a strategy to handle the human safety in a multi-robot scenario is devised. In the presented framework, it is foreseen that robots are in charge of performing any cooperative manipulation task which is parameterized by a proper task function. The devised architecture answers to the increasing demand of strict cooperation between humans and robots, since it equips a general multi-robot cell with the feature of making robots and human working together. The human safety is properly handled by defining a safety index which depends both on the relative position and velocity of the human operator and robots. Then, the multi-robot task trajectory is properly scaled in order to ensure that the human safety never falls below a given threshold which can be set in worst conditions according to a minimum allowed distance. Simulations results are presented in order to prove the effectiveness of the approach.},
	keywords = {Human-robot collaboration. Trajectory scaling. Multi-robot systems; Control and Systems Engineering},
	url = {https://www.sciencedirect.com/science/article/pii/S2405896318332464},
	doi = {10.1016/j.ifacol.2018.11.540},
	pages = {190--196},
	number = {22}
}

In this paper, a strategy to handle the human safety in a multi-robot scenario is devised. In the presented framework, it is foreseen that robots are in charge of performing any cooperative manipulation task which is parameterized by a proper task function. The devised architecture answers to the increasing demand of strict cooperation between humans and robots, since it equips a general multi-robot cell with the feature of making robots and human working together. The human safety is properly handled by defining a safety index which depends both on the relative position and velocity of the human operator and robots. Then, the multi-robot task trajectory is properly scaled in order to ensure that the human safety never falls below a given threshold which can be set in worst conditions according to a minimum allowed distance. Simulations results are presented in order to prove the effectiveness of the approach.

@article{
	11580_61356,
	author = {Arrichiello, Filippo and Marino, Alessandro and Pierri, Francesco},
	title = {Distributed Fault-Tolerant Control for Networked Robots in the Presence of Recoverable/Unrecoverable Faults and Reactive Behaviors},
	year = {2017},
	journal = {FRONTIERS IN ROBOTICS AND AI},
	volume = {4},
	abstract = {The paper presents an architecture for distributed control of multi-robot systems with an integrated fault detection, isolation, and recovery strategy. The proposed solution is based on a distributed observer-controller schema where each robot, by communicating only with its direct neighbors, is able to estimate the overall state of the system; such an estimate is then used by the controllers of each robot to achieve global missions as, for example, centroid and formation tracking. The information exchanged among the observers is also used to compute residual vectors that allow each robot to detect failures on anyone of the teammates, even if not in direct communication. The proposed strategy considers both recoverable and unrecoverable actuator faults as well as it properly manages the possible activation of reactive local control behaviors of the robots (e.g., the activation of obstacle avoidance strategy), which generate control inputs different from those required by the global mission control. In particular, when the robots are subject to recoverable faults, those are managed at a local level by computing a proper compensating control action. On the other side, when the robots are subject to unrecoverable faults, the faults are isolated from anyone of the teammates by means of a distributed fault detection and isolation strategy; then, the faulty robots are removed from the team and the mission is rearranged. The proposed strategy is validated via numerical simulations where the system properly identifies and manages the different cases of recoverable and unrecoverable actuator faults, as well as it manages the activation of local reactive control in an integrated case study.},
	url = {https://www.frontiersin.org/articles/10.3389/frobt.2017.00002/full},
	doi = {10.3389/frobt.2017.00002},
	pages = {1--12},
	number = {2}
}

The paper presents an architecture for distributed control of multi-robot systems with an integrated fault detection, isolation, and recovery strategy. The proposed solution is based on a distributed observer-controller schema where each robot, by communicating only with its direct neighbors, is able to estimate the overall state of the system; such an estimate is then used by the controllers of each robot to achieve global missions as, for example, centroid and formation tracking. The information exchanged among the observers is also used to compute residual vectors that allow each robot to detect failures on anyone of the teammates, even if not in direct communication. The proposed strategy considers both recoverable and unrecoverable actuator faults as well as it properly manages the possible activation of reactive local control behaviors of the robots (e.g., the activation of obstacle avoidance strategy), which generate control inputs different from those required by the global mission control. In particular, when the robots are subject to recoverable faults, those are managed at a local level by computing a proper compensating control action. On the other side, when the robots are subject to unrecoverable faults, the faults are isolated from anyone of the teammates by means of a distributed fault detection and isolation strategy; then, the faulty robots are removed from the team and the mission is rearranged. The proposed strategy is validated via numerical simulations where the system properly identifies and manages the different cases of recoverable and unrecoverable actuator faults, as well as it manages the activation of local reactive control in an integrated case study.

@conference{
	11580_71246,
	author = {Marino, Alessandro and Muscio, Giuseppe and Pierri, Francesco},
	title = {Distributed cooperative object parameter estimation and manipulation without explicit communication},
	year = {2017},
	booktitle = {ICRA 2017},
	abstract = {The paper presents a two stages distributed algorithm for cooperative manipulating an unknown object rigidly grasped by mobile manipulators, in the absence of both a central unit and any explicit information exchange among robots. In the first stage, robots cooperatively estimate the object kinematic and dynamic parameters by properly moving the object or applying specific contact wrenches. In the second stage, the estimated parameters are used in a distributed cooperative algorithm aimed at controlling the object pose while limiting both the squeezing wrenches exerted by the manipulators and the wrench exerted by the environment on the object. Numerical simulations demonstrate the feasibility of the approach.},
	keywords = {Distributed estimation; Distributed manipulation; Force control.},
	url = {http://ieeexplore.ieee.org/document/7989243/},
	doi = {10.1109/ICRA.2017.7989243},
	isbn = {978-1-5090-4633-1},	
	pages = {2110--2116}
}

The paper presents a two stages distributed algorithm for cooperative manipulating an unknown object rigidly grasped by mobile manipulators, in the absence of both a central unit and any explicit information exchange among robots. In the first stage, robots cooperatively estimate the object kinematic and dynamic parameters by properly moving the object or applying specific contact wrenches. In the second stage, the estimated parameters are used in a distributed cooperative algorithm aimed at controlling the object pose while limiting both the squeezing wrenches exerted by the manipulators and the wrench exerted by the environment on the object. Numerical simulations demonstrate the feasibility of the approach.

@conference{
	11580_71249,
	author = {Cirillo, Pasquale and Marino, Alessandro and Natale, Ciro and Di Marino, Emiliano and Chiacchio, Pasquale and De Maria, Giuseppe},
	title = {A low-cost and flexible solution for one-shot cooperative robotic drilling of aeronautic stack materials},
	year = {2017},
	publisher = {Elsevier B.V.},
	volume = {50},
	booktitle = {20th IFAC World Congress IFAC 2017},
	abstract = {In this paper, a general architecture for robotized cooperative drilling of hybrid metal/Carbon Fibre Reinforced Polymer (CFRP) aeronautic stacks is presented. The main objectives are to reduce non-added value operations as dismantling, deburring, cleaning and to improve hole quality by reducing the occurrence of delamination and inter-laminar cracking. The approach to achieve such targets is the adoption of a one-shot drilling automatized system based on low cost cooperative robots. The proposed technology consists in a general robot architecture and a cooperative drilling process using only standard low-cost robots and off-the shelf components. A first robot is in charge of drilling the hybrid stack, while a second manipulator ensures the right clamping force between the parts of the stack. Both robots are equipped with force control capabilities to control the generalized forces raising during the interaction with the stack. In this perspective, thanks to the adoption of a fuzzy inference system, the tuning of the force controllers might be carried out by operators that have knowledge of the drilling process but not of control system technology. Such feature makes the solution appealing for industrial production environments. The approach is experimentally validated on a cell with two Comau SmartSix robots drilling an aluminium-carbon fibre stack.},
	keywords = {Robotics technology; Robots manipulators; Control and Systems Engineering},
	url = {https://www.sciencedirect.com/science/article/pii/S240589631731491X},
	doi = {10.1016/j.ifacol.2017.08.1013},
	pages = {4602--4609},
	number = {1}
}

In this paper, a general architecture for robotized cooperative drilling of hybrid metal/Carbon Fibre Reinforced Polymer (CFRP) aeronautic stacks is presented. The main objectives are to reduce non-added value operations as dismantling, deburring, cleaning and to improve hole quality by reducing the occurrence of delamination and inter-laminar cracking. The approach to achieve such targets is the adoption of a one-shot drilling automatized system based on low cost cooperative robots. The proposed technology consists in a general robot architecture and a cooperative drilling process using only standard low-cost robots and off-the shelf components. A first robot is in charge of drilling the hybrid stack, while a second manipulator ensures the right clamping force between the parts of the stack. Both robots are equipped with force control capabilities to control the generalized forces raising during the interaction with the stack. In this perspective, thanks to the adoption of a fuzzy inference system, the tuning of the force controllers might be carried out by operators that have knowledge of the drilling process but not of control system technology. Such feature makes the solution appealing for industrial production environments. The approach is experimentally validated on a cell with two Comau SmartSix robots drilling an aluminium-carbon fibre stack.

@conference{
	11580_71243,
	author = {Marino, Alessandro},
	title = {A Decentralized Robust Adaptive Control for Tightly Connected Networked Lagrangian Systems},
	year = {2017},
	publisher = {IEEE},
	booktitle = {56th IEEE Conference on Decision and Control},
	abstract = {In this work, a decentralized control strategy for tightly
connected networked Lagrangian systems is designed. The main characteristics
of the solution is that it allows to both control the motion of
the handled object and the squeezing wrenches arising on it and this is
achieved by resorting to a layered architecture. At the top layer, robots
exploit consensus theory to distributedly estimate the full state of the
system and the object dynamics to estimate the squeezing wrenches,
while, at the second layer, a local adaptive control law is specified in
order to both control the local contribute to the squeezing wrenches
and the local motion of the robot. The effectiveness of the solution is
proven by employing 6-DOFs serial chain manipulators mounted on a
mobile platform to perform a cooperative load transportation task.},
	keywords = {Distributed observer-based control; cooperative manipulation},
	url = {https://ieeexplore.ieee.org/document/8264347/},
	doi = {10.1109/CDC.2017.8264347},
	isbn = {978-1-5090-2872-6},	
	pages = {4656--4661}
}

In this work, a decentralized control strategy for tightly connected networked Lagrangian systems is designed. The main characteristics of the solution is that it allows to both control the motion of the handled object and the squeezing wrenches arising on it and this is achieved by resorting to a layered architecture. At the top layer, robots exploit consensus theory to distributedly estimate the full state of the system and the object dynamics to estimate the squeezing wrenches, while, at the second layer, a local adaptive control law is specified in order to both control the local contribute to the squeezing wrenches and the local motion of the robot. The effectiveness of the solution is proven by employing 6-DOFs serial chain manipulators mounted on a mobile platform to perform a cooperative load transportation task.

@article{
	11580_71240,
	author = {Cirillo, Andrea and Cirillo, Pasquale and De Maria, Giuseppe and Marino, Alessandro and Natale, Ciro and Pirozzi, Salvatore},
	title = {Optimal custom design of both symmetric and unsymmetrical hexapod robots for aeronautics applications},
	year = {2017},
	journal = {ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING},
	volume = {44},
	abstract = {The Stewart parallel mechanism is used in various applications due to its high load-carrying capacity, accuracy and stiffness, such as flight simulation, spaceship aligning, radar and satellite antenna orientation, rehabilitation applications, parallel machine tools. However, the use of such parallel robots is not widespread due to three factors: the limited workspace, the singularity configurations existing inside the workspace, and the high cost. In this work, an approach to support the design of a cost-effective Stewart platform-based mechanism for specific applications and to facilitate the choice of suitable components (e.g., linear actuators and base and mobile plates) is presented. The optimal design proposed in this work has multiple objectives. In detail, it intends to maximize the payload and minimize the forces at each leg needed to counteract external forces applied to the mobile platform during positioning or manufacturing, or, in general, during specific applications. The approach also aims at avoiding reduction of the robot workspace through a kinematic optimization. Both symmetric and unsymmetrical geometries have been analysed to show how the optimal design approach can lead to effective results with different robot configurations. Moreover, these objectives are achieved through a dynamic optimization and several optimization algorithms were compared in terms of defined performance indexes.},
	keywords = {Aeronautics part positioningAdaptive fixtureStewart platformParallel robotOptimal designGenetic algorithm},
	url = {http://www.sciencedirect.com/science/article/pii/S0736584516302241},
	doi = {10.1016/j.rcim.2016.06.002},
	pages = {1--16}
}

The Stewart parallel mechanism is used in various applications due to its high load-carrying capacity, accuracy and stiffness, such as flight simulation, spaceship aligning, radar and satellite antenna orientation, rehabilitation applications, parallel machine tools. However, the use of such parallel robots is not widespread due to three factors: the limited workspace, the singularity configurations existing inside the workspace, and the high cost. In this work, an approach to support the design of a cost-effective Stewart platform-based mechanism for specific applications and to facilitate the choice of suitable components (e.g., linear actuators and base and mobile plates) is presented. The optimal design proposed in this work has multiple objectives. In detail, it intends to maximize the payload and minimize the forces at each leg needed to counteract external forces applied to the mobile platform during positioning or manufacturing, or, in general, during specific applications. The approach also aims at avoiding reduction of the robot workspace through a kinematic optimization. Both symmetric and unsymmetrical geometries have been analysed to show how the optimal design approach can lead to effective results with different robot configurations. Moreover, these objectives are achieved through a dynamic optimization and several optimization algorithms were compared in terms of defined performance indexes.

@conference{
	11580_71250,
	author = {Gasparri, Andrea and Marino, Alessandro},
	title = {A K-Hop Graph-Based Observer for Large-Scale Networked Systems},
	year = {2017},
	publisher = {IEEE},
	booktitle = {56th IEEE Conference on Decision and Control},
	abstract = {In this paper, we address the state estimation problem
for multi-agent systems interacting in large scale networks.
This research is motivated by the observation that in large-scale
networks for many practical applications and domains, each
agent only requires information concerning agents spatially close
to its location, let’s say topologically k-hop away. We propose a
scalable framework where each agent is able to estimate in finitetime
the state of its k-hop neighborhood by interacting only with
the agents belonging to its 1-hop neighborhood.},
	keywords = {Distributed observer},
	url = {https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8264361},
	doi = {10.1109/CDC.2017.8264361},
	isbn = {978-1-5090-2872-6},	
	pages = {4747--4752}
}

In this paper, we address the state estimation problem for multi-agent systems interacting in large scale networks. This research is motivated by the observation that in large-scale networks for many practical applications and domains, each agent only requires information concerning agents spatially close to its location, let’s say topologically k-hop away. We propose a scalable framework where each agent is able to estimate in finitetime the state of its k-hop neighborhood by interacting only with the agents belonging to its 1-hop neighborhood.

@article{
	11580_67049,
	author = {Marino, Alessandro and Pierri, Francesco and Arrichiello, Filippo},
	title = {Distributed Fault Detection Isolation and Accommodation for Homogeneous Networked Discrete-Time Linear Systems},
	year = {2017},
	journal = {IEEE TRANSACTIONS ON AUTOMATIC CONTROL},
	volume = {62},
	doi = {10.1109/TAC.2017.2694556},
	pages = {4840--4847},
	number = {9}
}

@conference{
	11580_71252,
	author = {Marino, Alessandro and Chiacchio, Pasquale},
	title = {Task-oriented decentralized adaptive control of cooperative manipulators},
	year = {2016},
	publisher = {IEEE},
	booktitle = {IEEE International Conference on Robotics and Biomimetics},
	abstract = {In this paper, the decentralized control in task
space of cooperative mobile arms with uncertain dynamics
is considered. The cooperative task is specified by means of
proper task-oriented variables depending on the end-effector
configurations of all the arms in the team. Therefore, because of
the lack of a centralized control unit, a two-layer architecture
is proposed. At the first level, each arm controller runs a
distributed observer in order to estimate the overall state of the
system. At the second level, this estimate is used to compute the
local control input in order to achieve the global cooperative
task. Moreover, the dynamic parameters of the arms might
not be perfectly known, thus the local control law has to be
made adaptive in order to counteract this uncertainty. The
approach was validated by simulation with 6-DOFs serial chain
manipulators mounted on a mobile platform and performing
a cooperative task.},
	keywords = {Networked Robots; Distributed control; Cooperative robots.},
	url = {http://ieeexplore.ieee.org/document/7866510/},
	doi = {10.1109/ROBIO.2016.7866510},
	isbn = {9781509043644},	
	pages = {1325--1330}
}

In this paper, the decentralized control in task space of cooperative mobile arms with uncertain dynamics is considered. The cooperative task is specified by means of proper task-oriented variables depending on the end-effector configurations of all the arms in the team. Therefore, because of the lack of a centralized control unit, a two-layer architecture is proposed. At the first level, each arm controller runs a distributed observer in order to estimate the overall state of the system. At the second level, this estimate is used to compute the local control input in order to achieve the global cooperative task. Moreover, the dynamic parameters of the arms might not be perfectly known, thus the local control law has to be made adaptive in order to counteract this uncertainty. The approach was validated by simulation with 6-DOFs serial chain manipulators mounted on a mobile platform and performing a cooperative task.

@conference{
	11580_71254,
	author = {Marino, Alessandro and Pierri, Francesco},
	title = {Discrete-time distributed state feedback control for multi-robot systems},
	year = {2016},
	publisher = {Institute of Electrical and Electronics Engineers Inc.},
	volume = {2016-},
	booktitle = {Proceedings - IEEE International Conference on Robotics and Automation},
	abstract = {In this paper, a general framework to control in a distributed way a system composed by multiple robots is proposed. Each robot is characterized by a discrete-time linear dynamics, and the whole system is controlled via a linear static feedback law with a feed-forward term. Usually, this form of the global control input requires a central unit or an all-to-all communication for computing the local control input of each robot. To counteract the lack of a central unit, each robot estimates, via a local observer, the overall state of the team, and such an estimate is used to compute its local control input as in the case a central unit was present. Two simulations case studies are provided in the framework of multi-robot optimal control and formation control.},
	keywords = {Software; Artificial Intelligence; Control and Systems Engineering; Electrical and Electronic Engineering},
	url = {http://ieeexplore.ieee.org/document/7487746/},
	doi = {10.1109/ICRA.2016.7487746},
	isbn = {9781467380263},	
	pages = {5350--5355}
}

In this paper, a general framework to control in a distributed way a system composed by multiple robots is proposed. Each robot is characterized by a discrete-time linear dynamics, and the whole system is controlled via a linear static feedback law with a feed-forward term. Usually, this form of the global control input requires a central unit or an all-to-all communication for computing the local control input of each robot. To counteract the lack of a central unit, each robot estimates, via a local observer, the overall state of the team, and such an estimate is used to compute its local control input as in the case a central unit was present. Two simulations case studies are provided in the framework of multi-robot optimal control and formation control.

@conference{
	11580_71248,
	author = {Marino, Alessandro and Cirillo, Pasquale and Natale, Ciro and Chiacchio, Pasquale and Pirozzi, Salvatore},
	title = {A general low-cost and flexible architecture for robotized drilling in aircraft assembly lines},
	year = {2016},
	publisher = {IEEE},
	booktitle = {2016 International Symposium on Power Electronics, Electrical Drives, Automation and Motion},
	abstract = {This paper describes a low-cost and flexible solution to automatize the drilling process in aircraft assembly
lines by using anthropomorphic robots. In nowadays aircraft
assembly lines, parts to be drilled (e.g., wing upper and lower
covers, spars, ribs, fuselage panels, etc.) are either drilled by
means of CNC machines or fixtures and jigs. In the latter case,
the drilling process is often performed by operators, and the
precision constraints are met thanks to reference drilling masks.
The devised solution allows to remove the human operator and
increase the flexibility of the system. At the same time, it is
suitable for existing assembly lines since it requires minimal
changes to them and is effective with almost every available
manipulators present on the market. To this aim, a well defined
procedure that leads a low-cost standard industrial manipulator
to meet drilling task requirements is presented. More in detail,
a standard arm manipulator carrying a commercial drilling end
effector and a force sensor is adopted. The effectiveness of the
designed architecture has been tested on a real setup equipped
with a Comau SmartSix manipulator in charge of drilling an
aluminum fuselage panel.},
	keywords = {Robotized drilling; Flexible Automation; Force
Control.},
	url = {http://ieeexplore.ieee.org/document/7525936/},
	doi = {10.1109/SPEEDAM.2016.7525936},
	isbn = {978-1-5090-2067-6},	
	pages = {1401--1408}
}

This paper describes a low-cost and flexible solution to automatize the drilling process in aircraft assembly lines by using anthropomorphic robots. In nowadays aircraft assembly lines, parts to be drilled (e.g., wing upper and lower covers, spars, ribs, fuselage panels, etc.) are either drilled by means of CNC machines or fixtures and jigs. In the latter case, the drilling process is often performed by operators, and the precision constraints are met thanks to reference drilling masks. The devised solution allows to remove the human operator and increase the flexibility of the system. At the same time, it is suitable for existing assembly lines since it requires minimal changes to them and is effective with almost every available manipulators present on the market. To this aim, a well defined procedure that leads a low-cost standard industrial manipulator to meet drilling task requirements is presented. More in detail, a standard arm manipulator carrying a commercial drilling end effector and a force sensor is adopted. The effectiveness of the designed architecture has been tested on a real setup equipped with a Comau SmartSix manipulator in charge of drilling an aluminum fuselage panel.

@conference{
	11580_71258,
	author = {Marino, Alessandro},
	title = {Distributed and fault tolerant control for a class of discrete time linear
systems},
	year = {2015},
	booktitle = {2015  Robotics: Science and Systems},
	keywords = {Distributed control; distributed fault detection},
	url = {http://mrs-rss2015.sciencesconf.org/},
	isbn = {978-0-9923747-1-6},	
	pages = {1--3}
}

@conference{
	11580_71257,
	author = {Marino, Alessandro and Pierri, Francesco},
	title = {Discrete-time distributed control and fault diagnosis for a class of linear systems},
	year = {2015},
	publisher = {IEEE},
	booktitle = {2015 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
	abstract = {This paper presents a solution to the problem of decentralized control, fault detection and isolation for teams of cooperative autonomous mobile vehicles. The strategy is carried out in the discrete time domain. A local observer is used by each agent to estimate the overall state of the team. This estimate is, then, used both to compute its local control input and isolate faulty teammates, even in absence of direct communication with them. For diagnosis purposes, a set of residual vectors, each of them sensible to a fault occurring on a single vehicle, is designed and an adaptive threshold is derived in order to avoid false alarms. The approach is validated via numerical simulations involving 4 vehicles moving in formation in a 3D environment.},
	keywords = {decentralised control, discrete time systems, distributed control, fault diagnosis, linear systems, mobile robots, multi-robot systems, observers},
	url = {http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=7353787},
	doi = {10.1109/IROS.2015.7353787},
	pages = {2974--2979}
}

This paper presents a solution to the problem of decentralized control, fault detection and isolation for teams of cooperative autonomous mobile vehicles. The strategy is carried out in the discrete time domain. A local observer is used by each agent to estimate the overall state of the team. This estimate is, then, used both to compute its local control input and isolate faulty teammates, even in absence of direct communication with them. For diagnosis purposes, a set of residual vectors, each of them sensible to a fault occurring on a single vehicle, is designed and an adaptive threshold is derived in order to avoid false alarms. The approach is validated via numerical simulations involving 4 vehicles moving in formation in a 3D environment.

@article{
	11580_63507,
	author = {Marino, Alessandro and Antonelli, Gianluca},
	title = {Experiments on sampling/patrolling with two Autonomous Underwater Vehicles},
	year = {2015},
	journal = {ROBOTICS AND AUTONOMOUS SYSTEMS},
	volume = {67},
	doi = {10.1016/j.robot.2014.09.030},
	pages = {61--71}
}

@conference{
	11580_55180,
	author = {Marino, A. and Pierri, F. and Chiacchio, P. and Chiaverini, Stefano},
	title = {Distributed fault detection and accommodation for a class of discrete-time linear systems},
	year = {2015},
	booktitle = {Proceedings of the 2015 IEEE International Conference on Information and Automation},
	abstract = {A decentralized fault diagnosis and accommodation scheme in discrete time for teams of cooperative mobile robots is presented. Each vehicle estimates, via a local observer, the overall state of the team, which is used to compute its local control input and a set of residual vectors. Each residual is sensible to a fault occurring on a single teammate, even not in direct communication, and the fault occurrence on a vehicle is declared if the corresponding residual exceeds a suitable adaptive threshold. Then, a recovery strategy, based on the estimate of the maximum detection time, is designed in order to remove the faulty teammates from the team and rearrange the mission. Numerical simulations, involving 5 vehicles moving in formation in a 3D environment, validate the proposed approach.},
	doi = {10.1109/ICInfA.2015.7279334},
	isbn = {978-1-4673-9104-7},	
	isbn = {978-1-4673-9104-7},	
	pages = {469--474}
}

A decentralized fault diagnosis and accommodation scheme in discrete time for teams of cooperative mobile robots is presented. Each vehicle estimates, via a local observer, the overall state of the team, which is used to compute its local control input and a set of residual vectors. Each residual is sensible to a fault occurring on a single teammate, even not in direct communication, and the fault occurrence on a vehicle is declared if the corresponding residual exceeds a suitable adaptive threshold. Then, a recovery strategy, based on the estimate of the maximum detection time, is designed in order to remove the faulty teammates from the team and rearrange the mission. Numerical simulations, involving 5 vehicles moving in formation in a 3D environment, validate the proposed approach.

@article{
	11580_52012,
	author = {Arrichiello, Filippo and Marino, Alessandro and Pierri, Francesco},
	title = {Observer-Based Decentralized Fault Detection and Isolation Strategy for Networked Multirobot Systems},
	year = {2015},
	journal = {IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY},
	volume = {23},
	abstract = {In this paper, we present a distributed fault detection and isolation (FDI) strategy for a team of networked robots that builds on a distributed controller-observer schema. Remarkably different from other works in literature, the proposed FDI approach makes each robot of the team able to detect and isolate faults occurring on other robots, even if they are not direct neighbors. By means of a local observer, each robot can estimate the overall state of the team and it can use such an estimate to compute its local control input to achieve global tasks. The same information used by the local observers is also used to compute residual vectors, whose aim is to allow the detection and the isolation of actuator faults occurring on any robot of the team. Adaptive thresholds are derived based on the dynamics of the residual vectors by considering the presence of nonzero initial observer estimation errors, and noise terms affecting state measurement and model dynamics. The approach is validated via both numerical simulations and experiments involving four Khepera III mobile robots.},
	doi = {10.1109/TCST.2014.2377175},
	pages = {1465--1476},
	number = {4}
}

In this paper, we present a distributed fault detection and isolation (FDI) strategy for a team of networked robots that builds on a distributed controller-observer schema. Remarkably different from other works in literature, the proposed FDI approach makes each robot of the team able to detect and isolate faults occurring on other robots, even if they are not direct neighbors. By means of a local observer, each robot can estimate the overall state of the team and it can use such an estimate to compute its local control input to achieve global tasks. The same information used by the local observers is also used to compute residual vectors, whose aim is to allow the detection and the isolation of actuator faults occurring on any robot of the team. Adaptive thresholds are derived based on the dynamics of the residual vectors by considering the presence of nonzero initial observer estimation errors, and noise terms affecting state measurement and model dynamics. The approach is validated via both numerical simulations and experiments involving four Khepera III mobile robots.

@article{
	11580_36306,
	author = {Marino, A. and Antonelli, Gianluca and Chiaverini, Stefano and Aguiar, A. P. and Pascoal, A.},
	title = {A decentralized strategy for multi-robot sampling/patrolling: theory and experiments},
	year = {2015},
	journal = {IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY},
	volume = {23},
	abstract = {This brief presents a decentralized coordinated strategy for multirobot patrolling missions. Patrolling is here interpreted within the framework of the sampling problem. To be applied in practice, several realistic constraints and the time/spatial variance of the information are explicitly considered. The proposed approach is well rooted in the concepts of Voronoi tessellations and Gaussian Processes. Each robot, based only on local information, computes the next point to visit according to a given performance criteria. Numerical simulations and experiments involving three autonomous marine surface robots in a harbor scenario at the Parque Expo site in Lisbon, Portugal, are presented.},
	keywords = {Gaussian Processes,multi-robot systems,patrolling,
surveillance.},
	doi = {10.1109/TCST.2014.2312550},
	pages = {313--322},
	number = {1}
}

This brief presents a decentralized coordinated strategy for multirobot patrolling missions. Patrolling is here interpreted within the framework of the sampling problem. To be applied in practice, several realistic constraints and the time/spatial variance of the information are explicitly considered. The proposed approach is well rooted in the concepts of Voronoi tessellations and Gaussian Processes. Each robot, based only on local information, computes the next point to visit according to a given performance criteria. Numerical simulations and experiments involving three autonomous marine surface robots in a harbor scenario at the Parque Expo site in Lisbon, Portugal, are presented.

@conference{
	11580_36881,
	author = {Arrichiello, Filippo and Marino, Alessandro and Pierri, F.},
	title = {Distributed fault-tolerant strategy for networked robots with both cooperative and reactive controls},
	year = {2014},
	publisher = {IEEE},
	address = {USA},
	booktitle = {Proceeding of the IEEE International Conference on Information and Automation},
	doi = {10.1109/ICInfA.2014.6932739},
	isbn = {978-1-4799-4100-1},	
	pages = {677--682}
}

@conference{
	11580_36882,
	author = {Arrichiello, Filippo and Marino, Alessandro and Pierri, F.},
	title = {Distributed fault detection and recovery for networked robots},
	year = {2014},
	publisher = {IEEE},
	address = {USA},
	booktitle = {Proceedings 2014 IEEE/RSJ International Conference on Intelligent Robots and Systems},
	isbn = {978-1-4799-6934-0},	
	pages = {3734--3739}
}

@conference{
	11580_45228,
	author = {Arrichiello, Filippo and Marino, A. and Pierri, F.},
	title = {A decentralized fault tolerant control strategy for multi-robot systems},
	year = {2014},
	publisher = {International Federation of Automatic Control},
	address = {Cape Town, South Africa},
	booktitle = {Proceedings of the 19th IFAC World Congress, 2014},
	doi = {10.3182/20140824-6-ZA-1003.01875},
	isbn = {978-3-902823-62-5},	
	pages = {6642--6647}
}

@article{
	11580_71241,
	author = {Basile, Francesco and Caccavale, Fabrizio and Chiacchio, Pasquale and Coppola, Jolanda and Marino, Alessandro and Gerbasio, Diego},
	title = {Automated synthesis of hybrid Petri net models for robotic cells in the aircraft industry},
	year = {2014},
	journal = {CONTROL ENGINEERING PRACTICE},
	volume = {31},
	abstract = {The use of robots in the aircraft structural assembly is a challenge. The presence of human operators, auxiliary systems and industrial robots makes hybrid the dynamic behavior of a robotic cell in this context. Here, the focus is on the automated synthesis of a model for the sequencing of the activities of a robotic cell in the aircraft industry. The cell model is obtained from a simple specification of resources and tasks, considered the main cell components, running the algorithm presented in this paper. The effectiveness of the model is shown using a case study defined by the ongoing European project LOCOMACHS (LOw COst Manufacturing and Assembly of Composite and Hybrid Structures, http://www.locomachs.eu/).},
	keywords = {Discrete event systems;Hybrid systems;Hybrid Petri nets;Manufacturing systems;Robotic workcells},
	url = {http://www.sciencedirect.com/science/article/pii/S0967066114001531},
	doi = {10.1016/j.conengprac.2014.05.008},
	pages = {35--49}
}

The use of robots in the aircraft structural assembly is a challenge. The presence of human operators, auxiliary systems and industrial robots makes hybrid the dynamic behavior of a robotic cell in this context. Here, the focus is on the automated synthesis of a model for the sequencing of the activities of a robotic cell in the aircraft industry. The cell model is obtained from a simple specification of resources and tasks, considered the main cell components, running the algorithm presented in this paper. The effectiveness of the model is shown using a case study defined by the ongoing European project LOCOMACHS (LOw COst Manufacturing and Assembly of Composite and Hybrid Structures, http://www.locomachs.eu/).

@inbook{
	11580_39857,
	author = {Antonelli, Gianluca and Arrichiello, Filippo and Casalino, G. and Chiaverini, Stefano and Marino, A. and Simetti, E. and Torelli, S.},
	title = {Harbour Protection Strategies with Multiple Autonomous Marine Vehicles},
	year = {2014},
	publisher = {Springer International Publishing},
	address = {Switzerland},
	booktitle = {Modelling and Simulation for Autonomous Systems},
	abstract = {This paper presents the ongoing research activities of the Italian Interuniversity Center of Integrated Systems for the Marine Environment, ISME, in the field of harbour protection with autonomous marine vehicles. In particular, two different strategies have been developed in the recent years and have been extensively tested both in numerical simulations and in scale experiments. In the first case, a set of vehicles is positioned around an asset to be protected on the base of an optimization process of two cost functions, namely, the maximization of minimum interception distance and the minimization of maximum interception time. When an intruder is detected, an on-line optimization process selects, among the different vehicles, the one that exhibits the lowest estimated time to the menace. A motion planning algorithm with real-time obstacle avoidance is then used to drive the vehicle toward the intruder. In the second case, a team of vehicles is required to dynamically patrol a certain region by means of a decentralized control approach. The proposed solution is based on the merging of two concepts, the Voronoi tessellations and the Gaussian processes, and it allows robustness with respect to events as temporary communication or vehicle losses. It also exhibits characteristics of flexibility/scalability with respect to the number of team-mates.},
	isbn = {978-3-319-13822-0},	
	pages = {241--261}
}

This paper presents the ongoing research activities of the Italian Interuniversity Center of Integrated Systems for the Marine Environment, ISME, in the field of harbour protection with autonomous marine vehicles. In particular, two different strategies have been developed in the recent years and have been extensively tested both in numerical simulations and in scale experiments. In the first case, a set of vehicles is positioned around an asset to be protected on the base of an optimization process of two cost functions, namely, the maximization of minimum interception distance and the minimization of maximum interception time. When an intruder is detected, an on-line optimization process selects, among the different vehicles, the one that exhibits the lowest estimated time to the menace. A motion planning algorithm with real-time obstacle avoidance is then used to drive the vehicle toward the intruder. In the second case, a team of vehicles is required to dynamically patrol a certain region by means of a decentralized control approach. The proposed solution is based on the merging of two concepts, the Voronoi tessellations and the Gaussian processes, and it allows robustness with respect to events as temporary communication or vehicle losses. It also exhibits characteristics of flexibility/scalability with respect to the number of team-mates.

@article{
	11580_36305,
	author = {Antonelli, Gianluca and Arrichiello, Filippo and Caccavale, F. and Marino, A.},
	title = {Decentralized time-varying formation control for multi-robot systems},
	year = {2014},
	journal = {THE INTERNATIONAL JOURNAL OF ROBOTICS RESEARCH},
	volume = {33},
	doi = {10.1177/0278364913519149},
	pages = {1029--1043},
	number = {7}
}

@conference{
	11580_36309,
	author = {Marino, A. and Antonelli, Gianluca},
	title = {Experimental Results of Coordinated Coverage by Autonomous Underwater Vehicles},
	year = {2013},
	publisher = {IEEE},
	address = {USA},
	booktitle = {Proceedings 2013 IEEE International Conference on Robotics and Automation},
	isbn = {9781467356404},	
	pages = {4126--4131}
}

@article{
	11580_71235,
	author = {Basile, Francesco and Caccavale, Fabrizio and Chiacchio, Pasquale and Coppola, Jolanda and Marino, Alessandro},
	title = {A decentralized kinematic control architecture for collaborative and cooperative multi-arm systems},
	year = {2013},
	journal = {MECHATRONICS},
	abstract = {In this paper a two-layer decentralized framework for kinematic control of cooperative and collaborative multi-robot systems is developed. The motion of the system is specified at the workpiece level, by adopting a task-oriented formulation for cooperative tasks. The first layer computes the motion of the single arms in the system. In detail, the control unit of each robot computes the end-effector motion references in a decentralized fashion on the basis of the knowledge of the assigned cooperative task and the motion references computed by its neighbors. Then, in the second layer, each control unit computes the reference joint motion of the corresponding manipulator from the end-effector reference motion. The approach is, then, tested in simulation on a work-cell composed by several manipulators, and experimentally on a dual-arm kinematically redundant work-cell composed by industrial manipulators.},
	keywords = {Cooperative manipulators; Industrial robotics; Task planning; Coordinated motion; Decentralized control},
	url = {http://www.sciencedirect.com/science/article/pii/S095741581300158X},
	doi = {10.1016/j.mechatronics.2013.08.008},
	pages = {1100--1112},
	number = {8}
}

In this paper a two-layer decentralized framework for kinematic control of cooperative and collaborative multi-robot systems is developed. The motion of the system is specified at the workpiece level, by adopting a task-oriented formulation for cooperative tasks. The first layer computes the motion of the single arms in the system. In detail, the control unit of each robot computes the end-effector motion references in a decentralized fashion on the basis of the knowledge of the assigned cooperative task and the motion references computed by its neighbors. Then, in the second layer, each control unit computes the reference joint motion of the corresponding manipulator from the end-effector reference motion. The approach is, then, tested in simulation on a work-cell composed by several manipulators, and experimentally on a dual-arm kinematically redundant work-cell composed by industrial manipulators.

@conference{
	11580_36310,
	author = {Antonelli, Gianluca and Arrichiello, Filippo and Caccavale, F. and Marino, A.},
	title = {Decentralized centroid and formation control for multi-robot systems},
	year = {2013},
	publisher = {IEEE},
	address = {Karlsruhe, D},
	booktitle = {Proceedings 2013 IEEE International Conference on Robotics and Automation},
	doi = {10.1109/ICRA.2013.6631068},
	isbn = {978-1-4673-5641-1},	
	pages = {3496--3501}
}

@article{
	11580_71237,
	author = {Caccavale, Fabrizio and Marino, Alessandro and Muscio, Giuseppe and Pierri, Francesco},
	title = {Discrete-Time Framework for Fault Diagnosis in Robotic Manipulators},
	year = {2013},
	journal = {IEEE TRANSACTIONS ON CONTROL SYSTEMS TECHNOLOGY},
	volume = {21},
	abstract = {In this paper, a discrete-time framework forbreak diagnosis of faults of joint sensors, wrist-mounted force/torque sensors, and actuators of robotic manipulators is devised. It is assumed that redundant joint sensor measurements are available. Sensor measurements, together with the estimates computed by two isolation observers, are processed by a decision-making system, providing detection and isolation of the faults of the joint sensors as well as healthy measurements. Then, healthy measurements are used to feed a bank of diagnostic observers aimed at detecting, isolating, and identifying faults of joint actuators and force/torque sensors. The framework is experimentally tested on a cooperative industrial setup, composed of two industrial robots with six degrees of freedom performing a cooperative task.},
	url = {http://ieeexplore.ieee.org/xpl/articleDetails.jsp?tp=&arnumber=6319596&queryText%3DDiscrete-Time+Framework+for+Fault+Diagnosis+in+Robotic+Manipulators},
	doi = {10.1109/TCST.2012.2212196},
	pages = {1858--1873},
	number = {5}
}

In this paper, a discrete-time framework forbreak diagnosis of faults of joint sensors, wrist-mounted force/torque sensors, and actuators of robotic manipulators is devised. It is assumed that redundant joint sensor measurements are available. Sensor measurements, together with the estimates computed by two isolation observers, are processed by a decision-making system, providing detection and isolation of the faults of the joint sensors as well as healthy measurements. Then, healthy measurements are used to feed a bank of diagnostic observers aimed at detecting, isolating, and identifying faults of joint actuators and force/torque sensors. The framework is experimentally tested on a cooperative industrial setup, composed of two industrial robots with six degrees of freedom performing a cooperative task.

@conference{
	11580_69646,
	author = {Marino, Alessandro and Antonelli, Gianluca},
	title = {Experimental results of coordinated sampling/patrolling by autonomous underwater vehicles},
	year = {2013},
	booktitle = {Proceedings 2013 IEEE International Conference on Robotics and Automation},
	doi = {10.1109/ICRA.2013.6631161},
	isbn = {978-1-4673-5643-5},	
	isbn = {978-1-4673-5641-1},	
	pages = {4141--4146}
}

@conference{
	11580_36880,
	author = {Arrichiello, Filippo and Marino, Alessandro and Pierri, F.},
	title = {A decentralized fault detection and isolation strategy for networked robots},
	year = {2013},
	publisher = {IEEE},
	address = {USA},
	booktitle = {Proceedings 2013 International Conference on Advanced Robotics},
	doi = {10.1109/ICAR.2013.6766562},
	isbn = {978-1-4799-2722-7},	
	pages = {1--6}
}

@conference{
	11580_36790,
	author = {Arrichiello, Filippo and Marino, A. and Meddahi, A.},
	title = {A decentralized observer for a general class of Lipschitz systems},
	year = {2013},
	publisher = {IEEE},
	address = {USA},
	booktitle = {Proceedings IEEE International Conference on Automation and Logistics},
	doi = {10.1109/ICInfA.2013.6720324},
	isbn = {978-1-4799-1334-3},	
	pages = {362--367}
}

@article{
	11580_23484,
	author = {Marino, A. and Parker, L. and Antonelli, Gianluca and Caccavale, F.},
	title = {A Decentralized Architecture for Multi-Robot Systems Based on the Null-Space-Behavioral Control with Application to Multi-Robot Border Patrolling},
	year = {2013},
	journal = {JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS},
	volume = {71},
	doi = {10.1007/s10846-012-9783-5},
	pages = {423--444},
	number = {3-4}
}

@article{
	11580_23483,
	author = {Antonelli, Gianluca and Arrichiello, Filippo and Caccavale, F. and Marino, A.},
	title = {A decentralized controller-observer scheme for multi-agent weighted centroid tracking},
	year = {2013},
	journal = {IEEE TRANSACTIONS ON AUTOMATIC CONTROL},
	volume = {58},
	abstract = {In this paper a decentralized controller-observer scheme for a multi-agent system is presented. The key idea is to develop, for each agent, an observer of the collective system's state and a motion controller. The observer is updated using only information from the agent itself and from its neighbors; the motion controller is designed in order to allow the team's weighted centroid to track an assigned time-varying reference. Convergence of the overall scheme is proven for directed and undirected communication graphs; moreover the extensions to the case of switching communication topologies and to the presence of saturation in the control input are discussed. Finally, numerical simulations are illustrated to validate the approach.},
	doi = {10.1109/TAC.2012.2220032},
	pages = {1310--1316},
	number = {5}
}

In this paper a decentralized controller-observer scheme for a multi-agent system is presented. The key idea is to develop, for each agent, an observer of the collective system's state and a motion controller. The observer is updated using only information from the agent itself and from its neighbors; the motion controller is designed in order to allow the team's weighted centroid to track an assigned time-varying reference. Convergence of the overall scheme is proven for directed and undirected communication graphs; moreover the extensions to the case of switching communication topologies and to the presence of saturation in the control input are discussed. Finally, numerical simulations are illustrated to validate the approach.

@conference{
	11580_20905,
	author = {Antonelli, Gianluca and Chiaverini, Stefano and Marino, A.},
	title = {A coordination strategy for multi-robot sampling of dynamic fields},
	year = {2012},
	publisher = {IEEE},
	address = {USA},
	booktitle = {IEEE International Conference on Robotics and Automation},
	doi = {10.1109/ICRA.2012.6224698},
	isbn = {9781612843803},	
	pages = {1113--1118}
}

@conference{
	11580_71251,
	author = {Munafò, Andrea and Calabrò, Vincenzo and Turetta, Alessio and Caffaz, Andrea and Marino, Alessandro and Caiti Giuseppe Casalino, Andrea and Indiveri, Giovanni and Antonelli, Gianluca},
	title = {Underwater robot networks: communication and cooperation},
	year = {2012},
	booktitle = {Proceedings IEEE Workshop on Underwater Communications Ucomms12},
	pages = {1--7}
}

@conference{
	11580_20907,
	author = {Antonelli, Gianluca and Arrichiello, Filippo and Caccavale, F. and Marino, A.},
	title = {A decentralized observer-controller scheme for centroid and formation control with bounded control input},
	year = {2012},
	publisher = {IFAC},
	address = {Santa Barbara, CA},
	booktitle = {3rd IFAC Workshop on Estimation and Control of Networked Systems},
	abstract = {In this paper a decentralized observer-controller scheme that allows to control the
centroid and the formation of multi-agent systems is presented. For each agent, the observer part
allows to estimate the state of the overall system, while the controller part is able to calculate
the input control for the given agent based on the estimated state of the system. The stability
analysis is carried out in the presence of input saturation and undirected/directed topologies.
Numerical simulations confirm the validity of the proposed approach.},
	url = {https://ac.els-cdn.com/S1474667015348424/1-s2.0-S1474667015348424-main.pdf?_tid=fb30139d-3cb0-4f3a-b2c6-a875648216e7&acdnat=1550494873_79ced157645b00b5a58b6527e8d4e2bc},
	doi = {10.3182/20120914-2-US-4030.00032},
	isbn = {9781622768837},	
	pages = {252--257},
	number = {26}
}

In this paper a decentralized observer-controller scheme that allows to control the centroid and the formation of multi-agent systems is presented. For each agent, the observer part allows to estimate the state of the overall system, while the controller part is able to calculate the input control for the given agent based on the estimated state of the system. The stability analysis is carried out in the presence of input saturation and undirected/directed topologies. Numerical simulations confirm the validity of the proposed approach.

@conference{
	11580_69645,
	author = {Marino, Alessandro and Antonelli, Gianluca and Aguiar, A. Pedro and Pascoal, Antonio},
	title = {A new approach to multi-robot harbour patrolling: Theory and experiments},
	year = {2012},
	booktitle = {2012 IEEE/RSJ International Conference on Intelligent Robots and Systems},
	doi = {10.1109/IROS.2012.6385864},
	isbn = {978-1-4673-1736-8},	
	isbn = {978-1-4673-1737-5},	
	isbn = {978-1-4673-1735-1},	
	pages = {1760--1765}
}

@conference{
	11580_23485,
	author = {Marino, A. and Antonelli, Gianluca and Aguiar, A. P. and Pascoal, A.},
	title = {Multi-robot harbor patrolling: a probabilistic approach},
	year = {2012},
	publisher = {IEEE/RSJ},
	address = {Vilamoura, PT},
	booktitle = {2012 IEEE/RSJ International Conference on Intelligent Robots and Systems},
	isbn = {978-1-4673-1737-5},	
	pages = {1760--1765}
}

@conference{
	11580_18399,
	author = {Antonelli, Gianluca and Arrichiello, Filippo and Marino, Alessandro and Caccavale, F.},
	title = {A decentralized controller-observer scheme for weighted centroid tracking},
	year = {2011},
	publisher = {IEEE},
	address = {USA},
	booktitle = {IEEE/RSJ International Conference on Intelligent Robots and Systems},
	doi = {10.1109/IROS.2011.6094400},
	isbn = {9781612844541},	
	pages = {2778--2783}
}

@article{
	11580_14073,
	author = {Antonelli, Gianluca and Curatella, C. and Marino, Alessandro},
	title = {Constrained motion planning for open-chain industrial robots},
	year = {2011},
	journal = {ROBOTICA},
	volume = {29},
	doi = {10.1017/S026357471000024X},
	pages = {403--420},
	number = {3}
}

@conference{
	11580_18398,
	author = {Antonelli, Gianluca and Marino, Alessandro and Chiaverini, Stefano},
	title = {Decentralized deployment with obstacle avoidance for AUVs},
	year = {2011},
	publisher = {International Federation of Automatic Control},
	address = {Milano, Italia},
	booktitle = {IFAC World Congress},
	abstract = {In multiple robotic system, the deployment problem consists in the optimal placement of the robots within a given environment. In this paper, a fully decentralized strategy, based on the so called Null-Space-based Behavioral control is proposed. The chosen behavioral approach allows to handle the eventual presence of obstacles in the environment in a task-priority fashion thus surmounting mono-task classical approaches. Numerical simulations confirm the validity of the proposed deployment strategy.},
	keywords = {Behaviou rpriority, autonomous mobile robots},
	url = {https://www.sciencedirect.com/science/article/pii/S1474667016456776},
	doi = {10.3182/20110828-6-IT-1002.02114},
	isbn = {9783902661937},	
	pages = {12807--12812},
	number = {1}
}

In multiple robotic system, the deployment problem consists in the optimal placement of the robots within a given environment. In this paper, a fully decentralized strategy, based on the so called Null-Space-based Behavioral control is proposed. The chosen behavioral approach allows to handle the eventual presence of obstacles in the environment in a task-priority fashion thus surmounting mono-task classical approaches. Numerical simulations confirm the validity of the proposed deployment strategy.

@inbook{
	11580_71242,
	author = {Pierri, Francesco and Marino, Alessandro and Muscio, Giuseppe and Caccavale, Fabrizio},
	title = {Research activities at the University of Basilicata},
	year = {2011},
	publisher = {CUES},
	booktitle = {Proceedings of the 1st PRISMA Workshop},
	isbn = {9788895028811},	
	pages = {109--131}
}

@article{
	11580_14072,
	author = {Antonelli, Gianluca and Caccavale, F. and Grossi, F. and Marino, Alessandro},
	title = {A Non-Iterative and effective procedure for simultaneous odometry and camera calibration for a differential drive mobile robot based on the singular value decomposition},
	year = {2010},
	journal = {INTELLIGENT SERVICE ROBOTICS},
	volume = {3},
	abstract = {Differential-drive mobile robots are usually
equipped with video cameras for navigation purposes.
In order to ensure proper operational capabilities of such
systems, several calibration steps are required to estimate the
video-camera intrinsic and extrinsic parameters, the relative
pose between the camera and the vehicle frame and the odometric parameters of the vehicle. In this paper, simultaneous
estimation of the aforementioned quantities is achieved by
a novel and effective calibration procedure. The proposed
calibration procedure needs only a proper set of landmarks,
on-board measurements given by the wheels encoders, and
the camera (i.e., a number of properly taken camera snapshots of the set of landmarks). A major advantage of the
proposed technique is that the robot is not required to follow
a specific path: the vehicle is asked to roughly move around
the landmarks and acquire at least three snapshots at some
approximatively known configurations. Moreover, since the
whole calibration procedure does not use external measurement devices, it can be used to calibrate, on-site, a team of
mobile robots with respect to the same inertial frame, given
by the position of the landmarks’ tool. Finally, the proposed
algorithm is systematic and does not require any iterative step. Numerical simulations and experimental results,
obtained by using a mobile robot Khepera III equipped with
a low-cost camera, confirm the effectiveness of the proposed
technique.},
	keywords = {Camera calibration; Odometry calibration; Mobile robots},
	doi = {10.1007/s11370-010-0067-2},
	pages = {163--173},
	number = {3}
}

Differential-drive mobile robots are usually equipped with video cameras for navigation purposes. In order to ensure proper operational capabilities of such systems, several calibration steps are required to estimate the video-camera intrinsic and extrinsic parameters, the relative pose between the camera and the vehicle frame and the odometric parameters of the vehicle. In this paper, simultaneous estimation of the aforementioned quantities is achieved by a novel and effective calibration procedure. The proposed calibration procedure needs only a proper set of landmarks, on-board measurements given by the wheels encoders, and the camera (i.e., a number of properly taken camera snapshots of the set of landmarks). A major advantage of the proposed technique is that the robot is not required to follow a specific path: the vehicle is asked to roughly move around the landmarks and acquire at least three snapshots at some approximatively known configurations. Moreover, since the whole calibration procedure does not use external measurement devices, it can be used to calibrate, on-site, a team of mobile robots with respect to the same inertial frame, given by the position of the landmarks’ tool. Finally, the proposed algorithm is systematic and does not require any iterative step. Numerical simulations and experimental results, obtained by using a mobile robot Khepera III equipped with a low-cost camera, confirm the effectiveness of the proposed technique.

@conference{
	11580_16804,
	author = {Antonelli, Gianluca and Marino, Alessandro},
	title = {Smooth 3-Dimensional Path Generation with Guaranteed Maximum Distance from Via-Points},
	year = {2010},
	publisher = {International Federation of Automatic Control/Elsevier},
	address = {Lecce, I},
	booktitle = {The 7th IFAC Symposium on Intelligent Autonomous Vehicles},
	doi = {10.3182/20100906-3-IT-2019.00053},
	isbn = {9783902661876},	
	pages = {----},
	number = {1}
}

@conference{
	11580_16803,
	author = {Indiveri, G. and Antonelli, Gianluca and Caiti, A. and Casalino, G. and Birk, A. and Pascoal, A. and Caffaz, A. and Marino, Alessandro},
	title = {The CO3AUVs (Cooperative Cognitive Control for Autonomous Underwater Vehicles) Project: overview and current progresses},
	year = {2010},
	publisher = {International Federation of Automatic Control/Elsevier},
	address = {Lecce, I},
	booktitle = {The 7th IFAC Symposium on Intelligent Autonomous Vehicles},
	doi = {10.3182/20100906-3-IT-2019.00053},
	isbn = {9783902661876},	
	pages = {----},
	number = {1}
}

@conference{
	11580_71256,
	author = {Caccavale, F. and Marino, A. and Pierri, F.},
	title = {Sensor fault diagnosis for manipulators performing interaction tasks},
	year = {2010},
	publisher = {IEEE},
	booktitle = {IEEE International Symposium on  Industrial Electronics (ISIE), 2010},
	abstract = {In this paper a fault diagnosis approach for sensor faults in cooperative robotic manipulators involved in interaction tasks is presented. The approach is developed for a two-arm cooperative workcell, although it can be easily applied to single-manipulator systems. A bank of discrete-time model-based diagnostic observers is adopted to detect, isolate and identify failures of both sensors at the joints of the manipulators and force/torque sensors mounted at the wrists. The effectiveness of the proposed scheme has been experimentally tested on a cooperative industrial setup composed by two six-dof COMAU Smart-3 S robots.},
	keywords = {discrete time systems, failure analysis, fault diagnosis, force control, force sensors, industrial robots, manipulators, multi-robot systems, observers, torque control},
	url = {http://ieeexplore.ieee.org/abstract/document/5637775/},
	doi = {10.1109/ISIE.2010.5637775},
	isbn = {978-142446391-6},	
	pages = {2121--2126}
}

In this paper a fault diagnosis approach for sensor faults in cooperative robotic manipulators involved in interaction tasks is presented. The approach is developed for a two-arm cooperative workcell, although it can be easily applied to single-manipulator systems. A bank of discrete-time model-based diagnostic observers is adopted to detect, isolate and identify failures of both sensors at the joints of the manipulators and force/torque sensors mounted at the wrists. The effectiveness of the proposed scheme has been experimentally tested on a cooperative industrial setup composed by two six-dof COMAU Smart-3 S robots.

@conference{
	11580_14074,
	author = {Antonelli, Gianluca and Caccavale, F. and Grossi, F. and Marino, Alessandro},
	title = {Simultaneous Calibration of Odometry and Camera for a Differential Drive Mobile Robot},
	year = {2010},
	publisher = {IEEE Press},
	address = {Anchorage, AK},
	booktitle = {Proceedings 2010 IEEE International Conference on Robotics and Automation},
	doi = {10.1109/ROBOT.2010.5509954},
	isbn = {9781424450404},	
	pages = {5417--5422}
}

@conference{
	11580_1463,
	author = {Antonelli, Gianluca and Chiaverini, Stefano and Curatella, C. and Marino, A.},
	title = {Constrained Motion Planning for Industrial Robots},
	year = {2009},
	booktitle = {Proceedings 2009 IEEE International Conference on Automation and Logistics},
	abstract = {In this paper, a motion planning algorithm for robot manipulators is presented. Due to physical constraints, the robot manipulators are prone to several limitations such as, e.g., the maximum torque at joints; in addition, task-oriented constraints are given by, e.g., the requirement to respect a given path even in presence of saturations. The algorithm developed approaches the motion planning algorithm from a wide perspective, solving the joint as well as the Cartesian motion, both for the point-to-point and the fly movements, according to several, severe, constraints. An extensive testing phase, both via numerical simulation and experiments, using the current constraints, has been performed leading to excellent results.},
	url = {https://ieeexplore.ieee.org/document/5262622},
	doi = {10.1109/ICAL.2009.5262622},
	pages = {1934--1939}
}

In this paper, a motion planning algorithm for robot manipulators is presented. Due to physical constraints, the robot manipulators are prone to several limitations such as, e.g., the maximum torque at joints; in addition, task-oriented constraints are given by, e.g., the requirement to respect a given path even in presence of saturations. The algorithm developed approaches the motion planning algorithm from a wide perspective, solving the joint as well as the Cartesian motion, both for the point-to-point and the fly movements, according to several, severe, constraints. An extensive testing phase, both via numerical simulation and experiments, using the current constraints, has been performed leading to excellent results.

@conference{
	11580_1464,
	author = {Marino, A. and Parker, L. and Antonelli, Gianluca and Caccavale, F.},
	title = {Fuzzy Behavioral Control for Multi-Robot Border Patrol},
	year = {2009},
	booktitle = {17th Mediterranean Conference on Control and Automation},
	abstract = {This paper deals with the problem of multi-robot border patrolling. The patrolling algorithm is designed by resorting to the behavioral control framework and is organized in a hierarchical structure. Several Elementary Behaviors are defined, which are the basis of the concept of Action, placed at a higher level of abstraction with respect to the behaviors. Each Action is obtained by properly combining multiple Elementary Behaviors via the Null-Space-Behavioral control framework. For the sake of robustness, the overall patrolling algorithm is fully decentralized, since explicit communication between robots is not needed. A a Fuzzy Inference System is designed to select the proper Action according to local sensor information only. The algorithm has been validated in simulation as well as experimentally on a setup composed by three Pioneer robots.},
	keywords = {Behavioral control; Border Patrol; Platoon of vehicles; Multi-robot systems; Swarm Robotics.},
	url = {https://ieeexplore.ieee.org/document/5164547},
	doi = {10.1109/MED.2009.5164547},
	pages = {246--251}
}

This paper deals with the problem of multi-robot border patrolling. The patrolling algorithm is designed by resorting to the behavioral control framework and is organized in a hierarchical structure. Several Elementary Behaviors are defined, which are the basis of the concept of Action, placed at a higher level of abstraction with respect to the behaviors. Each Action is obtained by properly combining multiple Elementary Behaviors via the Null-Space-Behavioral control framework. For the sake of robustness, the overall patrolling algorithm is fully decentralized, since explicit communication between robots is not needed. A a Fuzzy Inference System is designed to select the proper Action according to local sensor information only. The algorithm has been validated in simulation as well as experimentally on a setup composed by three Pioneer robots.

@conference{
	11580_13464,
	author = {Marino, Alessandro and Parker, L. and Antonelli, Gianluca and Caccavale, F.},
	title = {Behavioral Control for Multi-Robot Perimeter Patrol: A Finite State Automata approach},
	year = {2009},
	booktitle = {Proceedings 2009 IEEE International Conference on Robotics and Automation},
	doi = {10.1109/ROBOT.2009.5152710},
	pages = {831--836}
}

@conference{
	11580_3156,
	author = {Marino, A and Parker, L. E. and Antonelli, Gianluca and Caccavale, F and Chiaverini, Stefano},
	title = {A fault-tolerant modular control approach to multi-robot perimeter patrol},
	year = {2009},
	publisher = {IEEE},
	address = {USA},
	booktitle = {Proceedings 2009 IEEE International Conference on Robotics and Biomimetics (ROBIO 2009)},
	doi = {10.1109/ROBIO.2009.5420581},
	isbn = {9781424447749},	
	isbn = {9781424447756},	
	pages = {735--740}
}

@conference{
	11580_3099,
	author = {Antonelli, Gianluca and Chiaverini, Stefano and Curatella, C and Marino, A.},
	title = {Prioritized Closed-Loop Inverse Kinematic Algorithms for Redundant Robotic Systems with Velocity Saturations},
	year = {2009},
	publisher = {IEEE},
	address = {USA},
	booktitle = {Proceedings 2009 IEEE International Conference on Automation and Logistics (ICAL 2009)},
	doi = {10.1109/ICAL.2009.5262622},
	isbn = {9781424438037},	
	isbn = {9781424438044},	
	pages = {1934--1939}
}

@article{
	11580_71238,
	author = {Caccavale, F. and Marino, A and Chiacchio, P and Villani, L},
	title = {Experiments of Impedance Control for a Dual-Arm Cooperative System},
	year = {2008},
	journal = {AUTOMAZIONE E STRUMENTAZIONE},
	volume = {56},
	pages = {82--89}
}

@article{
	11580_71236,
	author = {Caccavale, F and Chiacchio, P. and Marino, A and Villani, L},
	title = {Six-DOF Impedance Control of Dual-Arm Cooperative Manipulators},
	year = {2008},
	journal = {IEEE/ASME TRANSACTIONS ON MECHATRONICS},
	volume = {13},
	abstract = {In this paper, the problem of impedance control of dual-arm cooperative manipulators is studied. A general impedance control scheme is adopted, which encompasses a centralized impedance control strategy, aimed at conferring a compliant behavior at the object level, and a decentralized impedance control, enforced at the end-effector level, aimed at avoiding large internal loading of the object. Remarkably, the mechanical impedance behavior is defined in terms of geometrically consistent stiffness. The overall control scheme is based on a two-loop arrangement, where a simple proportional integral derivative inner motion loop is adopted for each manipulator, while an outer loop, using force and moment measurements at the robots wrists, is aimed at imposing the desired impedance behaviors. The developed control scheme is experimentally tested on a dual-arm setup composed of two 6-DOF industrial manipulators carrying a common object. The experimental investigation concerns the four different controller configurations that can be achieved by activating/deactivating the single impedance controllers.},
	keywords = {Control; cooperative manipulators; robotics},
	doi = {10.1109/TMECH.2008.2002816},
	pages = {576--586},
	number = {5}
}

In this paper, the problem of impedance control of dual-arm cooperative manipulators is studied. A general impedance control scheme is adopted, which encompasses a centralized impedance control strategy, aimed at conferring a compliant behavior at the object level, and a decentralized impedance control, enforced at the end-effector level, aimed at avoiding large internal loading of the object. Remarkably, the mechanical impedance behavior is defined in terms of geometrically consistent stiffness. The overall control scheme is based on a two-loop arrangement, where a simple proportional integral derivative inner motion loop is adopted for each manipulator, while an outer loop, using force and moment measurements at the robots wrists, is aimed at imposing the desired impedance behaviors. The developed control scheme is experimentally tested on a dual-arm setup composed of two 6-DOF industrial manipulators carrying a common object. The experimental investigation concerns the four different controller configurations that can be achieved by activating/deactivating the single impedance controllers.

@conference{
	11580_71247,
	author = {Caccavale, F. and Chiacchio, P. and DE SANTIS, A. and Marino, A. and Villani, L.},
	title = {An Experimental Investigation on Impedance Control for Dual-Arm Cooperative Systems},
	year = {2007},
	publisher = {Institute of Electrical and Electronics Engineers ( IEEE )},
	booktitle = {2007 IEEE/ASME International Conference on Advanced Intelligent Mechatronics},
	abstract = {In this paper, an experimental comparison between different impedance control approaches for cooperative manipulators is presented. The experiments have been performed on a dual-arm setup composed by two six-dof industrial manipulators carrying a common object. A general impedance control scheme is adopted, which encompasses a centralized impedance control strategy aimed at conferring a compliant behavior at the object level, and a decentralized impedance control, enforced at the end-effector level, aimed at avoiding large internal loading of the object. The scheme allows individual activation/deactivation of each impedance strategy. The experimental investigation concerns the four different controller configurations that can be achieved by activating/deactivating the single impedance controllers.},
	keywords = {Industrial robotics , impedance control , force feedback},
	doi = {10.1109/AIM.2007.4412434},
	pages = {180--185}
}

In this paper, an experimental comparison between different impedance control approaches for cooperative manipulators is presented. The experiments have been performed on a dual-arm setup composed by two six-dof industrial manipulators carrying a common object. A general impedance control scheme is adopted, which encompasses a centralized impedance control strategy aimed at conferring a compliant behavior at the object level, and a decentralized impedance control, enforced at the end-effector level, aimed at avoiding large internal loading of the object. The scheme allows individual activation/deactivation of each impedance strategy. The experimental investigation concerns the four different controller configurations that can be achieved by activating/deactivating the single impedance controllers.