Complex models for locomotion of biped humanoids robots, that include balance and dynamics, have been developed in the last decades. In parallel, complex methods for grasping have also been developed, and in particular, the H2T lab has a lot of experience in grasping and planning manipulation tasks for upper body humanoids robots.
However, the underlying mathematical models that describe both a humanoid and a hand grasping an object are similar. The exploration of such similarities and the transfer of methodologies between both fields is promising path to fully exploit the already proposed solutions as well as to find new ones.
Different topics to explore such parallelisms include, but are not limited to, in-hand manipulation vs. whole-body motions, grasp synthesis vs. step planning, balance conditions vs. stable grasps, etc.