SUMMARY

The field of soft robotics promises systems that adapt to humans or environments by utilizing lightweight, compliant structures instead of relying solely on sophisticated control strategies. While this morphological approach to adaptation has been shown effective, soft robotic systems are still quite limited in practical applications due to challenges in developing compatible soft sensors, complexity in modeling such systems, and limitations in soft actuator bandwidth and force output capabilities.

The proposed work seeks to improve the controllability of soft-actuator-powered systems, specifically to enable high-bandwidth dynamic capabilities required in assisted human running and jumping, and for propulsion of human-scale legged mobile robots. A threefold approach will be taken, focusing on pneumatic artificial muscles – soft actuators that contract under pressure. The first objective is to develop a pneumatic muscle design with integrated sensing for kinematic and dynamic state estimation of muscle-actuated robots without sacrificing the muscle’s compliant nature. The second objective is to develop a design optimization strategy to maximize a robot’s dynamic capabilities given the force-length profile and bandwidth limitations of pneumatic muscles. The third objective is to develop a simple on/off pressure control approach to achieve more sophisticated pre-planned robot motion sequences via trajectory optimization using a combination of analytical and data-driven models.

https://bluejeans.com/496368980