The goal was to demonstrate that the hand mechanism works as long as something can pull on the tendons. We used Dynamixel servos (with Maxon motors and good gear boxes inside) but you are right, ideally we would use something intrinsically compliant. The only practical technology that can do that right now are…