Justin E. Seipel

Postdoctoral Fellow

Lab phone: 510-643-5183
Cell phone: 609-647-7972
Fax: 510-643-6264

Email: jseipel@berkeley.edu

Animals are capable of performing amazing locomotion behaviors in complex environments. To understand how animals move we need to understand how the control system and morphological-structural systems of an organism interact with the environment to create a whole organism-environment system with complex, adaptive properties that can produce relatively simple and useful (directed) behavior like locomotion. Further, characterizing and quantifying how properties of the whole system affect the ‘fitness' of an organism can suggest principles for the optimality of locomotion in the context of evolutionary cost functions and designed robotic systems. For example, quantifying the affect of morphology on the stability and robustness of gaits, while varying control architecture and environment, can help us compare the functionality of different forms under different conditions, thus also potentially enabling comparative study of otherwise far-removed species.

Dynamical systems theory is the natural background on which to work out the interaction of control, morphology, and environment in organismal systems. Analyzable mathematical models that are representative of the whole organism and environment can help us develop a deeper intuition for complex organismal systems and can simultaneously provide us with computationally inexpensive prediction tools. Since blind simplification can often leave out important features, it is evident that we need to perform experiments and run detailed simulations along with analyzing simple models to ultimately find and understand the essential features, properties, and principles of a given system and its behaviors. The principles of motion discovered may enable robot design and the development of simple models for prediction, planning, and real-time control of robotic systems.