Improvement of distributed control algorithms for robots carrying an object

We consider the problem of moving a pair of omnidirectional robots carrying a ladder between start and goal positions in an obstacle-free plane as quickly as possible, using a distributed algorithm in which each robot decides its motion individually. In this paper we propose a modification to an existing algorithm for this problem and examine the performance of the resulting algorithm by computer simulation. The simulation results indicate that the resulting algorithm performs better in terms of time than previous algorithms in certain situations, and nearly as well in most other cases.