Dedale: Demonstrating a Realistic Testbed for Decentralized Multi-agents Problems
- 48 Downloads
The Dedale platform is a peer-to-peer multi-agent testbed dedicated to the study of MAS coordination, learning and decision-making problems under realistic hypotheses: Asynchrony, partial observability, uncertainty, heterogeneity, open environments, limited communication and computation. Dedale facilitates the implementation of reproducible and realistic experiments in discrete and (3D) continuous environments. Agents can face cooperative or competitive exploration, patrolling, pickup and delivery, treasure(s) or agent(s) hunt problems with teams of dozens of heterogeneous agents. This paper presents the demonstration elaborated in order to exibit the platform’s capabilities.
KeywordsAgent testbed Coordination Learning Decision-making
- 1.Adil, K., Jiang, F., Liu, S., Jifara, W., Tian, Z., Fu, Y.: State-of-the-art and open challenges in rts game-ai and starcraft. Int. J. Adv. Comput. Sci. Appl. 8(12), 16–24 (2017)Google Scholar
- 3.Chevaleyre, Y.: Theoretical analysis of the multi-agent patrolling problem. In: Proceedings of the IEEE/WIC/ACM - IAT, pp. 302–308. IEEE (2004)Google Scholar
- 4.Hausknecht, M., Mupparaju, P., Subramanian, S., Kalyanakrishnan, S., Stone, P.: Half field offense: an environment for multiagent learning and ad hoc teamwork. In: AAMAS Adaptive Learning Agents (ALA) Workshop (2016)Google Scholar
- 5.Resnick, C., et al.: Pommerman: a multi-agent playground. arXiv:1809.07124 (2018)