Task Allocation in a Multi-Robot System Using RRT* Algorithm

Task allocation in multi-robot systems is essential to optimize system efficiency, particularly in scenarios where multiple robots can achieve tasks common to all the robots. This paper presents a hybrid approach to independent non-constrained task allocation utilizing the RRT* algorithm. The allocation process is based on a cost function calculated when the RRT* algorithm finds a path from initial robot configuration to goal node. Path planning is essential in task allocation and to generate feasible paths in robot environments, especially if obstacles are scattered around the task-space. The RRT* algorithm is capable of finding optimal paths in cluttered environments, incorporating modifications for improved convergence and efficiency. The study evaluates the proposed approach using a multi-robot system comprising two ABB IRB1410 manipulators. The system demonstrates the effectiveness of the algorithm in allocating tasks based on their cost and achievability. Experiments are conducted to showcase the algorithm's ability to navigate obstacles and optimize task allocation. Overall, the paper contributes to task allocation strategies in multi-robot systems by offering a practical implementation using RRT* capability to minimize the cost function of the task.