Learning Policies for Dynamic Coalition Formation in Multi-Robot Task Allocation

On IEEE Xplore and on arXiv

Abstract We propose a decentralized, learning-based framework for dynamic coalition formation in Multi-Robot Task Allocation (MRTA). Our approach extends MAPPO by integrating spatial action maps, robot motion planning, intention sharing, and task allocation revision to enable effective and adaptive coalition formation. Extensive simulation studies confirm the effectiveness of our model, enabling each robot to rely solely on local information to learn timely revisions of task selections and form coalitions with other robots to complete collaborative tasks. The results also highlight the proposed framework’s ability to handle large robot populations and adapt to scenarios with diverse task sets.

Citation

@article{Bezerra2025,
   author = {Lucas C. D. Bezerra and Ataide M.G. Dos Santos and Shinkyu Park},
   doi = {10.1109/LRA.2025.3592080},
   issn = {23773766},
   issue = {9},
   journal = {IEEE Robotics and Automation Letters},
   keywords = {Multi-robot systems, cooperating robots, reinforcement learning},
   pages = {9216-9223},
   publisher = {Institute of Electrical and Electronics Engineers Inc.},
   title = {Learning Policies for Dynamic Coalition Formation in Multi-Robot Task Allocation},
   volume = {10},
   url = {https://ieeexplore.ieee.org/document/11091462},
   year = {2025}
}