We are mainly interested in understanding how the passive dynamics of the body and the environment can be used by an agent like a robot to simplify what it has to compute to maintain stability and an optimum information exchange with the environment. For instance, when we are required to estimate the weight of a ball, we toss it up and down several times with slightly varying joint stiffnesses. When we are required to probe a soft tissue using a finger, people use various finger stiffness control strategies during probing. Do these examples mean that the central nervous system modulates the impedance of joints to regulate how it wants to feel the World in addition to using stiffness control to regulate the extent to which active control is enforced on the environment?
All our work in different projects in the lab tries to address this general theme known as morphological computation. We use various paradigms like human robot interaction through soft and hard reins, robotic palpation of soft tissue using controllable stiffness joints, and passive dynamic walking on soft terrains. We very often build our own simple robotic platforms to test various specific phenomena. See more at Thrishlab