Objects Relocation in Clutter with Robot Manipulators via Tree-based Q-Learning Algorithm: Analysis and Experiments. Golluccio, G., Di Lillo, P., Di Vito, D., Marino, A., & Antonelli, G. JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS, 2022. ![Objects Relocation in Clutter with Robot Manipulators via Tree-based Q-Learning Algorithm: Analysis and Experiments [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex 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_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}
}
Downloads: 0
{"_id":"mjJzkLwg2qkxbQXuj","bibbaseid":"golluccio-dilillo-divito-marino-antonelli-objectsrelocationinclutterwithrobotmanipulatorsviatreebasedqlearningalgorithmanalysisandexperiments-2022","author_short":["Golluccio, G.","Di Lillo, P.","Di Vito, D.","Marino, A.","Antonelli, G."],"bibdata":{"bibtype":"article","type":"article","author":[{"propositions":[],"lastnames":["Golluccio"],"firstnames":["G."],"suffixes":[]},{"propositions":[],"lastnames":["Di","Lillo"],"firstnames":["P."],"suffixes":[]},{"propositions":[],"lastnames":["Di","Vito"],"firstnames":["D."],"suffixes":[]},{"propositions":[],"lastnames":["Marino"],"firstnames":["A."],"suffixes":[]},{"propositions":[],"lastnames":["Antonelli"],"firstnames":["G."],"suffixes":[]}],"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","bibtex":"@article{\n\t11580_93947,\n\tauthor = {Golluccio, G. and Di Lillo, P. and Di Vito, D. and Marino, A. and Antonelli, G.},\n\ttitle = {Objects Relocation in Clutter with Robot Manipulators via Tree-based Q-Learning Algorithm: Analysis and Experiments},\n\tyear = {2022},\n\tjournal = {JOURNAL OF INTELLIGENT & ROBOTIC SYSTEMS},\n\tvolume = {106},\n\tabstract = {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.},\n\tkeywords = {Motion planning; Reinforcement learning; Task planning},\n\turl = {https://link.springer.com/article/10.1007/s10846-022-01719-9},\n\tdoi = {10.1007/s10846-022-01719-9},\n\tnumber = {2}\n}\n","author_short":["Golluccio, G.","Di Lillo, P.","Di Vito, D.","Marino, A.","Antonelli, G."],"key":"11580_93947","id":"11580_93947","bibbaseid":"golluccio-dilillo-divito-marino-antonelli-objectsrelocationinclutterwithrobotmanipulatorsviatreebasedqlearningalgorithmanalysisandexperiments-2022","role":"author","urls":{"Paper":"https://link.springer.com/article/10.1007/s10846-022-01719-9"},"keyword":["Motion planning; Reinforcement learning; Task planning"],"metadata":{"authorlinks":{}}},"bibtype":"article","biburl":"https://bibbase.org/network/files/LhpEBBMAufCZxdRLG","dataSources":["a9jNc2NyyC7d58MTq","YjpAZwNicNHANadbW","9HNP8ouNgGHQfHxfN"],"keywords":["motion planning; reinforcement learning; task planning"],"search_terms":["objects","relocation","clutter","robot","manipulators","via","tree","based","learning","algorithm","analysis","experiments","golluccio","di lillo","di vito","marino","antonelli"],"title":"Objects Relocation in Clutter with Robot Manipulators via Tree-based Q-Learning Algorithm: Analysis and Experiments","year":2022}