BioLoco Research

IVALab BioLocomotion Research

Our research is geared towards mid-level control of biologically-inspired robotic systems. We consider low-level control to live at the level of the gait, and the design of gaits that preserve locomotion stability of the robot itself. This is a pretty big research problem for legged locomotion, since legged robots can fall. However, for other biologically-inspired robots, stability is a bit easier to achieve when moving over planar ground. We focus on this class of robotic systems, with the aim of using known gaits to derive controlled equations of motion that are analytically tractable. This means that it is possible to do one of two things: (1) make steps towards proving that a control policy will work or (2) transform the controlled gait equations into control equations that make the robotic animal look like a traditional robot (like a Roomba or an automobile). Once its mathematically looks like a traditional robot, we can use all of the nice machinery associated to typical robots for achieving autonomous operation.

So far, we have been studying snake locomotion with the intent to figure out how to perform planning, trajectory synthesis, and feedback control for this class of bio-inspired robots. We got the theory mostly worked out for flat planar surface, and hope that we can experimentally verify the mathematics. So far, we've got rectilinear motion down, and are going to take care of sidewinding next. Lateral undulation is a bit trickier since the motion dynamics are a bit more complex.

Investigators Involved

Collaborators