Exact Approaches to the Multi-agent Collective Construction Problem
- 685 Downloads
The multi-agent collective construction problem tasks agents to construct any given three-dimensional structure on a grid by repositioning blocks. Agents are required to also use the blocks to build ramps in order to access the higher levels necessary to construct the building, and then remove the ramps upon completion of the building. This paper presents a mixed integer linear programming model and a constraint programming model of the problem, either of which can exactly optimize the problem, as previous efforts have only considered heuristic approaches. The two models are evaluated on several small instances with a large number of agents. The plans clearly show the swarm behavior of the agents. The mixed integer linear programming model is able to find optimal solutions faster than the constraint programming model and even some existing incomplete methods due to its highly-exploitable network flow substructures.
KeywordsClassical planning Multi-agent planning Multi-agent path finding Blocksworld Swarm robotics
The research at the University of Southern California was supported by the National Science Foundation (NSF) under grant numbers 1724392, 1409987, 1817189, 1837779, and 1935712.
- 1.Cai, T., Zhang, D., Kumar, T.K.S., Koenig, S., Ayanian, N.: Local search on trees and a framework for automated construction using multiple identical robots. In: Proceedings of the International Conference on Autonomous Agents and Multi-Agent Systems (2016)Google Scholar
- 3.Jones, C., Mataric, M.: Automatic synthesis of communication-based coordinated multi-robot systems. In: Proceedings of the IEEE/RSJ International Conference on Intelligent Robots and Systems (2004)Google Scholar
- 4.Koenig, S., Kumar, T.K.S.: A case for collaborative construction as testbed for cooperative multi-agent planning. In: Proceedings of the ICAPS-2017 Scheduling and Planning Applications Workshop (2017)Google Scholar
- 5.Kumar, T.K.S., Jung, S., Koenig, S.: A tree-based algorithm for construction robots. In: Proceedings of the International Conference on Automated Planning and Scheduling (2014)Google Scholar
- 7.Napp, N., Klavins, E.: Robust by composition: programs for multi-robot systems. In: Proceedings of the IEEE International Conference on Robotics and Automation (2010)Google Scholar
- 8.Petersen, K., Nagpal, R., Werfel, J.: TERMES: an autonomous robotic system for three-dimensional collective construction. In: Proceedings of Robotics: Science and Systems (2011)Google Scholar
- 9.Sartoretti, G., Wu, Y., Paivine, W., Kumar, T.K.S., Koenig, S., Choset, H.: Distributed reinforcement learning for multi-robot decentralized collective construction. In: Correll, N., Schwager, M., Otte, M. (eds.) Distributed Autonomous Robotic Systems. SPAR, vol. 9, pp. 35–49. Springer, Cham (2019). https://doi.org/10.1007/978-3-030-05816-6_3CrossRefGoogle Scholar
- 10.Stern, R., et al.: Multi-agent pathfinding: definitions, variants, and benchmarks. In: Proceedings of the Symposium on Combinatorial Search (2019)Google Scholar
- 11.Vaidyanathan, B., Ahuja, R.K.: Minimum cost flows. In: Wiley Encyclopedia of Operations Research and Management Science. Wiley (2011)Google Scholar