Cyber-Enabled Exercise Machines

The broad goal of this project is to advance both the foundations and the enabling technologies in the field of human-machine systems, with a focus on exercise and rehabilitation machines.

One of the motivations of our project is the observation that advances in exercise devices are mostly seen in their haptic interfaces rather than their intrinsic capabalities. Also, our collaboration with NASA on exercise machines for astronauts has motivaed us to investigate this aspect of human-machine interaction.

Our working hypothesis is that physical training for athletic conditioning, rehabilitation and special environments can be improved by endowing machines with optimality-seeking adaptive behavior.

Specifically, we seek to lay the foundations for the design, optimization and real-time control of machines which vary their mechanical impedance in pursuit of an optimal objective related to exercise performance. 

To this effect, our team is performing research in the areas of human performance and exercise science, biomechanical modeling and predictive simulation, estimation theory, control theory and mechatronic systems.

Our first year has produced advances in each of the above areas, as elaborated in the poster and video.