Key points are not available for this paper at this time.
Policy gradient methods can be used for mechanical and computational co-design of robot manipulators.
Chen et al. (Wed,) studied this question.