Abstract
•The proposed Inter-Star algorithm used for inter-reconfiguring robot path plans.•The theoretical algorithm time complexity is proven by empirical data.•The algorithm shows significant improvement in computing time performance.•The algorithm is validated with up to 3-robot path plans for inter-reconfiguration.
Reconfigurable robots can overcome physical limitations by combining them for larger strength and area coverage and splitting them to fit spatial constraints. The robustness and versatility of such robots are difficult to appreciate without some level of autonomy built into such systems. The paper identified that one of the issues encountered in this development is the scalability of such autonomous systems. To address one of the many problems in this issue, this paper proposes a novel approach to modifying the popularly used Astar algorithm and enabling scalability in computing path plans during the combining and splitting phase of inter-reconfiguration. The proposed Inter-Star algorithm achieves its time complexity O(logn·x) for multi-robot path planning in converging and diverging path plans. The team has studied the inefficiencies and redundancies of iterating the algorithm as applied widely by many contributors and practitioners. The tested and proposed algorithms space and time complexity are analyzed and discussed. At a small scale, the results of the modification have been shown to improve the scalability by at least 3.02 times more scalable than the original Astar algorithm. In addition, up to 3 Wasp Biggie robots, an inter-reconfigurable robot platform, are used to show the performance of this algorithm against Dijkstra’s and Astar algorithms. In both simulations and experiments, the proposed solution has shown superior computational time performances.