\n \n \n
\n
\n\n \n \n \n \n \n Computer Vision for safety in Collaborative Assembly.\n \n \n \n\n\n \n David Kötter.\n\n\n \n\n\n\n Master's thesis, RWTH Aachen University, September 2022.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@thesis{koetter2022,\n author = {K{\\"o}tter, David},\n title = {Computer Vision for safety in Collaborative Assembly},\n year = {2022},\n month = sep,\n school = {{RWTH Aachen University}},\n type = {mathesis},\n advisor = {Behery, Mohamed},\n keywords = {Robotic Assembly, Human-Robot Collaboration, Behavior Trees}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n Verifying Belief Program by Symbolic Dynamic Programming.\n \n \n \n\n\n \n Huang Qinfei.\n\n\n \n\n\n\n Master's thesis, RWTH Aachen University, Aachen, 2022.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n \n \n abstract \n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@thesis{ qinfei_msthesis,\n title = {Verifying Belief Program by\nSymbolic Dynamic Programming},\n author = {Qinfei, Huang},\n year = {2022},\n school = {RWTH Aachen University},\n type = {mathesis},\n address = {Aachen},\n advisor = {Daxin, Liu},\n keywords = {program verification, epistemic program, probabilistic program},\n abstract = {\nThe action programming language Golog is a powerful tool for expressing\nhigh-level agent behaviors. One advantage of programming in Golog is that its logical foundation, the situation calculus, makes formal analyse of programs’ behaviour\npossible. As a probabilistic extension of the Golog programming language family,\nBelle and Levesque proposed a framework called belief programs where every\naction and sensing could be noisy and test refers to the agent’s subjective belief. These\ncharacteristics, among others, make belief programs rather suitable for robot control\nin a partial-observable uncertain environment.\nBefore deployment, it is desirable to verify programs meet certain properties as\ndesired. To fulfil the task, Liu and Lakemeyer reconsider the proposal of belief\nprograms by Belle and Levesque based on a probabilistic modal logic of belief\nand actions, i.e. the logic DSp. Among other things, their proposal allows expressing\nprogram properties in a variant of PCTL logic. Liu and Lakemeyer [Liuar] also show\nthat the verification problem is closely related to model-checking infinite horizon partial\nobservable Markov decision process (POMDP), therefore is undecidable. They show\nthat the problem is strongly undecidable even under very restrictive settings, nevertheless, they prove that for bounded temporal properties, the verification problem could\nbe decidable, assuming the underlying logic is decidable.\nDespite many theoretical results they achieved, a general algorithm that actually\nconducts the verification is missing. In this thesis, we propose an algorithm based\non symbolic dynamic programming to verify belief programs. Symbolic dynamic programming (SDP) is a generalization of the dynamic programming technique for solving\nMarkov decision processes (MDPs) that exploits the symbolic structure in the solution\nof relational and first-order logical MDPs through a lifted version of dynamic programming. Such technique suits it well in verifying belief programs.\n}\n}\n\n\n
\n
\n\n\n
\n The action programming language Golog is a powerful tool for expressing high-level agent behaviors. One advantage of programming in Golog is that its logical foundation, the situation calculus, makes formal analyse of programs’ behaviour possible. As a probabilistic extension of the Golog programming language family, Belle and Levesque proposed a framework called belief programs where every action and sensing could be noisy and test refers to the agent’s subjective belief. These characteristics, among others, make belief programs rather suitable for robot control in a partial-observable uncertain environment. Before deployment, it is desirable to verify programs meet certain properties as desired. To fulfil the task, Liu and Lakemeyer reconsider the proposal of belief programs by Belle and Levesque based on a probabilistic modal logic of belief and actions, i.e. the logic DSp. Among other things, their proposal allows expressing program properties in a variant of PCTL logic. Liu and Lakemeyer [Liuar] also show that the verification problem is closely related to model-checking infinite horizon partial observable Markov decision process (POMDP), therefore is undecidable. They show that the problem is strongly undecidable even under very restrictive settings, nevertheless, they prove that for bounded temporal properties, the verification problem could be decidable, assuming the underlying logic is decidable. Despite many theoretical results they achieved, a general algorithm that actually conducts the verification is missing. In this thesis, we propose an algorithm based on symbolic dynamic programming to verify belief programs. Symbolic dynamic programming (SDP) is a generalization of the dynamic programming technique for solving Markov decision processes (MDPs) that exploits the symbolic structure in the solution of relational and first-order logical MDPs through a lifted version of dynamic programming. Such technique suits it well in verifying belief programs. \n
\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Multi-Agent System for Production Processes in the Automotive Industry.\n \n \n \n \n\n\n \n Maximilian Kerpen.\n\n\n \n\n\n\n Bachelor's thesis, RWTH Aachen University, April 2022.\n
\n\n
\n\n
\n\n
\n\n \n \n Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 28 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@thesis{Kerpen2022,\n author = {Maximilian Kerpen},\n title = {Multi-Agent System for Production Processes in the Automotive Industry},\n url = {https://kbsg.rwth-aachen.de/theses/kerpen2022.pdf},\n year = {2022},\n month = apr,\n school = {{RWTH Aachen University}},\n type = {bathesis},\n advisor = {Liebenberg, Martin and Viehmann, Tarik},\n keywords = {MAS, Agent, Ontologies, Distributed Systems, WWL}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Utilizing a PDDL Goal Model for Goal Reasoning with the CLIPS Executive.\n \n \n \n \n\n\n \n Patrick Lusts.\n\n\n \n\n\n\n Bachelor's thesis, RWTH Aachen University, April 2022.\n
\n\n
\n\n
\n\n
\n\n \n \n Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 12 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@thesis{Lusts2022,\n author = {Patrick Lusts},\n title = {Utilizing a PDDL Goal Model for Goal Reasoning with the CLIPS Executive},\n url = {https://kbsg.rwth-aachen.de/theses/lusts2022.pdf},\n year = {2022},\n month = apr,\n school = {{RWTH Aachen University}},\n type = {bathesis},\n advisor = {Viehmann, Tarik},\n keywords = {Goal Reasoning, PDDL, Planning, Clips Executive}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Whole-Body Manipulation on Mobile Robots Using Parallel Position-Based Visual Servoing.\n \n \n \n \n\n\n \n Matteo Tschesche.\n\n\n \n\n\n\n Master's thesis, RWTH Aachen University, May 2022.\n
\n\n
\n\n
\n\n
\n\n \n \n Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 29 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@thesis{Tschesche2022,\n author = {Matteo Tschesche},\n title = {Whole-Body Manipulation on Mobile Robots Using Parallel Position-Based Visual Servoing},\n url = {https://kbsg.rwth-aachen.de/theses/tschesche2022.pdf},\n year = {2022},\n month = may,\n school = {{RWTH Aachen University}},\n type = {mathesis},\n advisor = {Hofmann, Till},\n keywords = {visual servoing, robotics, manipulation, RCLL}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n An Approach to Instance Segmentation and Pose Estimation for Automated Dismantling of Lithium-Ion Batteries for High-Quality Recycling.\n \n \n \n\n\n \n Anna-Maria Meer.\n\n\n \n\n\n\n Master's thesis, RWTH Aachen University, July 2022.\n
\n\n
\n\n
\n\n
\n\n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n\n \n \n \n \n \n \n \n\n \n \n \n\n\n\n
\n
@thesis{Meer2022,\n author = {Anna-Maria Meer},\n title = {An Approach to Instance Segmentation and Pose Estimation for Automated Dismantling of Lithium-Ion Batteries for High-Quality Recycling},\n year = {2022},\n month = jul,\n school = {{RWTH Aachen University}},\n type = {mathesis},\n advisor = {Christina Ionescu and Till Hofmann and Alexander von Rohr}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n Transfer Reinforcement Learning in a Distributed Laser-Based Manufactoring System.\n \n \n \n \n\n\n \n Irina Sutyrina.\n\n\n \n\n\n\n Master's thesis, RWTH Aachen University, July 2022.\n
\n\n
\n\n
\n\n
\n\n \n \n Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 14 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@thesis{Sutyrina2022,\n author = {Irina Sutyrina},\n title = {Transfer Reinforcement Learning in a Distributed Laser-Based Manufactoring System},\n url = {https://kbsg.rwth-aachen.de/theses/sutyrina2022.pdf},\n year = {2022},\n month = jul,\n school = {{RWTH Aachen University}},\n type = {mathesis},\n advisor = {Viehmann, Tarik and Kröger, Moritz},\n keywords = {reinforcement learning, transfer learning, laser-based manufactoring}\n}\n\n
\n
\n\n\n\n
\n\n\n
\n
\n\n \n \n \n \n \n \n A CLIPS-Based CBR Framework for Pass Schedule Recommendations in Open-Die Forging.\n \n \n \n \n\n\n \n Sarah Soomro.\n\n\n \n\n\n\n Bachelor's thesis, RWTH Aachen University, November 2022.\n
\n\n
\n\n
\n\n
\n\n \n \n Paper\n \n \n\n \n\n \n link\n \n \n\n bibtex\n \n\n \n\n \n \n \n 9 downloads\n \n \n\n \n \n \n \n \n \n \n\n \n \n \n \n \n \n \n \n \n\n\n\n
\n
@thesis{Soomro2022,\n author = {Sarah Soomro},\n title = {A CLIPS-Based CBR Framework for Pass Schedule Recommendations in Open-Die Forging},\n url = {https://kbsg.rwth-aachen.de/theses/soomro2022.pdf},\n year = {2022},\n month = nov,\n school = {{RWTH Aachen University}},\n type = {bathesis},\n advisor = {Viehmann, Tarik},\n keywords = {case-based reasoning, pass schedule recommendations, open-die forging}\n}\n\n
\n
\n\n\n\n
\n\n\n\n\n\n