Simon Sponberg

Graduate Student

Lab phone: 510-643-5183
Fax: 510-643-6264
Email: sponberg@berkeley.edu

I am a doctoral candidate studying the neuromechanical control of stability and maneuverability, mostly in cockroaches. Legged animals show remarkable locomotor abilities, many unmatched by current human designs. They accomplish these behaviors in unpredictable environments from which they must extract relevant sensory information and negotiate unpredictable perturbations. Using techniques from biomechanics and neurobiology I seek to develop and test hypotheses about the structure and function of control strategies used by locomoting animals. These hypotheses are driven from a dynamic systems and control theoretical view of locomotion and in turn inform mathematical and physical (robotic) models of locomotor principles.

My current research goals are to use detailed mechanistic studies on model organisms within a broader comparative framework to reveal general principles along three lines. How is sensory feedback (e.g. reflexes) integrated with feedforward control (e.g. CPGs) and the physical dynamics of a springy, damping, viscous musclo-skeletal system? How do these integrated control systems enable performance, particularly stability and maneuverability? How is the myriad available sensory information processed and filtered to provide information relevant for a particular behavioral task?

My dissertation research focuses on control of preferred speed running in cockroaches. In investigating questions of stability, I specifically look at cockroaches encountering repeated random perturbations and highly repeatable, predefined single perturbations while recording neural activation of putative control muscles and measuring body dynamics. With respect to maneuverability, I specifically look at the tactile wall following navigation exhibited by cockroaches with long antennae. Finally in trying to understand the coupling of neural feedback in a mechanically tuned system, I simulate the effects of specific reflexes by directly rewriting muscle activation patterns so that I can systematically explore muscles’ control potential and investigate the context (behavioral and mechanical) dependency of their effects. Throughout, I collaborate with engineers to transfer useful biomechanical concepts to legged robotics and to use robots as physical models for understanding biology.

In the future, I will continue to use the techniques that I have developed during my dissertation (functional electrical stimulation, kinematics analysis, accelerometry, perturbation analysis, etc) but will apply them to new systems and behaviors while incorporate more neuroethological approaches. I am particularly interested in aspects of motor coding and have begun to develop techniques similar to quantifying receptive fields in sensory neurophysiology. I am seeking a postdoc with a theoretical, experimental, and/or applied neuroscience component to compliment my understanding of the biomechanics of locomotion and sensorimotor integration. Additionally, the approach I have taken for my dissertation is driven primarily from a dynamic systems perspective and I seek to probe the same questions from more of an information theoretic perspective, as the two approaches need better theoretical and experimental integration.