Toward Parallel Task Allocation in Multi-Robot Environment Based on Auction Mechanism with Deadlock Constrains

Md Sabbir Hossain, Naoki Fukuta

Abstract: In this paper we present an approach for parallel task allocation in multi-robot environments based on an auction mechanism with deadlock constraints. By using the Iterative Combinatorial Auction (ICA) mechanism and integrating Proportional-Integral (PI) algorithms, our propose approach offers a robust and efficient solution to address critical challenges in task allocation. Throughout the paper, we will highlight the importance of effective task allocation in multi-robot systems and identified the potential risks associated with deadlock situations. By incorporating deadlock constraints into the auction mechanism and dynamically adjusting bidding strategies using PI algorithms, our approach would ensures continuous task execution while maximizing system performance.