An extension of the plan-merging paradigm for multi-robot coordination

This paper reports on our efforts to extend the plan-merging paradigm, in order to improve its ability to adapt to various execution contexts. They concern the algorithmic as well as the architectural issues. The scheme can be parameterized, allowing to choose and even to update online robot priorities, and to take into account various constraints imposed by the application. These lead to different coordination policies. The architectural improvements allowed a better distribution of the computation load, leading to more complex applications, for instance, the coordination of multiple mobile manipulators. The result is a more flexible scheme that should widen its application fields. We describe the main ingredients of this new scheme and illustrate its implementation through two examples.