› Research : GRASP :: Reconfigurable Robot
›› Passive Reconfigurable Robot
Self reconfigurable robot forms require a global geometrical structure from a variable number of homogenous, macroscale independent modules. They can rearrange to new morphologies autonomously and act together to accomplish diverse tasks, such as economic mass production of units, graceful degradation of function when damaged (also called Self-repair), and the ability to transition into topologies suitable to the task at hand. These individual units have onboard power and intelligence, and are assumed to exist in a Brownian motion. For these modules to passively and stochastically reconfigure, we deduced graph grammar rules for a particular configuration and then implemented the algorithm onboard based on these learned rules. There have been lots of successful work on deliberate detereministic planning, but our goal was to study a novel approach to self organization that is based on a passive, stochastic reconfiguration.
››› Poster [PDF]
››› Report [PDF]