Multi-Head Path Planning of SwarmItFIX Agents: A Markov Decision Process Approach
SwarmItFIX is a flexible fixturing system designed and developed for sheet metal manufacturing applications. In this work, an artificial intelligence technique developed based on Markov Decision Process (MDP) is implemented for the task of multi-head based fixturing suitable for operations such as drilling, milling etc. The trajectory of the machining tool is split into various line segments for the ease of applying the MDP. The MDP is applied for obtaining the state parameters such as position coordinates and orientation of the intermediate heads of all the segments individually. The state parameters of the corner heads are calculated directly by finding the intersection angles of consecutive segments. In MDP, the evolution of utility values of all states are obtained using value iteration algorithm. For the convergence of the optimal policy of all segments in the contour, the policy iteration algorithm is employed. The execution time of both the algorithms were observed. Finally, the multi-head path planning model has been developed based on the algorithm having least execution time. The model returns the optimal policies for all the segments. Computer simulations are performed, and the results show that the multi-robot heads could be positioned in an effective manner in the given trajectory. Therefore, in near future the developed multi-head path planning model will be tested and implemented into the SwarmItFIX setup at the PMAR laboratory, University of Genoa.
KeywordsSwarmItFIX Multi-head path planning Markov decision process Policy iteration Value iteration
Unable to display preview. Download preview PDF.
This work is an outcome of the sponsored project (Project No: MEC/DIMEUG-ITALY/2015-16/MSK/008/02) supported by the University of Genoa, Italy.
- 1.Leonardo, Luis., Zoppi, Matteo., Xiong, Li., Zlatanov, Dimiter., Molfino, Rezia.: SwarmItFIX: a multi-robot-based reconfigurable fixture. Industrial Robot: An International Journal 40 (4), 320-328 (2013).Google Scholar
- 3.Li, Xiong., Zoppi, Matteo., de Leonardo, Luis., Molfino, Rezia.: Adaptable fixturing heads for swarm fixtures: discussion of two designs. Proceedings of the ASME 2012 11th Biennial Conference on Engineering Systems Design and Analysis (ESDA2012). 4, 99-106, (2012).Google Scholar
- 8.Sagar, Keerthi., Zlatanov, Dimiter., Zoppi, Matteo., Nattero, Cristiano., Muthuswamy, Sreekumar.: Multi-goal path planning for robotic agents with discrete-step locomotion. Proceedings of ASME 2017 International Design Engineering Technical Conferences & Computers and Information in Engineering Conference IDETC, Ohio, USA (2017).Google Scholar
- 9.Sagar, Keerthi., Zlatanov, Dimiter., Zoppi, Matteo., Nattero, Cristiano., Muthuswamy, Sreekumar.: Orientation planning for multi-agents with discrete-step locomotion and multiple goals. Proceedings of ASME 2018 International Design Engineering Technical Conferences & Computers and Information in Engineering Conference IDETC, Quebec, Canada (2018).Google Scholar
- 11.Stuart Russell., Peter Norvig.: Artificial intelligence: a modern approach. 3rd edn. Pearson Education, India (2017).Google Scholar
- 12.J, Burlet., O, Aycard., T, Fraichard.: Robust motion planning using markov decision processes and quadtree decomposition. Proceedings of IEEE International Conference on Robotics and Automation 2004. 2820-2825. LA, USA, (2004).Google Scholar
- 13.Zoppi, M., Molfino, R., and Zlatanov, D.: Bench and method for the support and manufacturing of parts with complex geometry, US Patent 8495811 (2013).Google Scholar
- 14.Rezia, Molfino., Matteo, Zoppi., Dimiter, Zlatanov.: Reconfigurable swarm fixtures. ASME/IFToMM International Conference on Reconfigurable Mechanisms and Robots (ReMAR), (2009).Google Scholar