Core Concepts
The author proposes a framework for multi-robot systems to adapt to faults and disturbances through epistemic planning, enabling efficient task allocation in communication-restricted environments.
Abstract
The content discusses a novel framework for multi-robot systems, focusing on task allocation with intermittent interactions. It introduces epistemic planning to address faults and disturbances, improving system efficiency. The proposed approach is validated through simulations and real-world experiments, showcasing significant improvements over baseline heuristics.
The paper emphasizes the importance of cooperative planning in multi-robot systems, especially when faced with limited communication or unexpected events. By integrating epistemic logic and Monte Carlo tree search, the framework enables robots to adapt dynamically to changing conditions. Simulations demonstrate the effectiveness of the proposed method in various scenarios, outperforming traditional approaches.
Furthermore, laboratory experiments validate the framework's practical application using Bitcraze Crazyflies in a controlled environment. The results highlight the system's ability to handle faults, replan tasks efficiently, and complete missions successfully. Future research directions include optimizing computation time and expanding strategies for more complex environments.
Stats
"The square environment used for simulations has dimensions of 30 × 30, 30 × 30, 90 × 90, and 150 × 150 [m]."
"Each robot propagates three particles in the proposed approach."
"The initial maximum speed of each vehicle is 5 [m/s], with second and third particles traveling at linear speeds of 80% and 60% of the maximum speed."
"Maximum communication range is set at 5 [m] from the center of each robot."