Towards an optimal avoidance strategy for collaborative robots