That is such a vague question. You're asking if some type of material that gains charge due to deformation can replicate muscles for a robotic arm. There is no answer to that unless there's some kind of setup that you're just replacing the material with a ceramic. Or am I missing something?
I don't think piezoelectric devices would provide sufficient displacement+force to be of any practical use in robotic devices. The work density is too low. Look at all of the various devices being developed by DARPA...they are still constructed with umbilicals pumping hydraulic fluids to provide sufficient power to move links fast enough, and with enough force.