About Us

Our research over the past 12 years has developed through three main phases. In the early part, we studied biologically inspired sensors and sensory computation systems, primarily in the form of vision sensors. Typically, these systems were implemented with Very Large Scale Integrated (VLSI) technology and were used to extract information about the environment and to guide the “attention” of other computation systems. In the middle part, we studied how these systems can be hosted onto robots. At that point we also started to model spinal neural circuits in silicon, and develop robots to study legged locomotion. We developed a close collaboration with Prof. M. A. Lewis of the University of Arizona and Iguana Robotics, Inc. Both VLSI circuits and biomorphic robots were developed and used in these studies. (We define a biomorphic robot as a mechanical device and control system that mimics the form and function of its living counterpart. For example, legged robots and prosthetic limbs are biomorphic robots.) More recently, we have evolved this work to include brain-machine interfaces and neural prosthesis devices. Specifically, we have started looking at spinal and cortical prosthesis devices and robotic systems to restore function after injury and for human augmentation. This new area has required close collaboration with neuroscientists to gain access to animal models (i.e. lamprey and cat preparations). Our recent work has included various experiments to understand neurophysiology of spinal neural circuits, to interface with them, to decode their sensory-motor relationships, and to use these relationships to control biomorphic robots. We plan to continue to expand this area research, while leveraging my laboratory’s expertise in VLSI circuits and systems, visual and acoustical information processing, neuromorphic computation systems and biomorphic robotics. Some of our contributions over the years are:

  • Analog Neural Computer
  • Visual Motion Estimation and Tracking
  • Focal-Plane Spatial Processing and the Computation-On-Readout (COR) Architecture
  • Legged Locomotion and Central Pattern Generation
  • Spinal Neuroprosthesis Devices and the Control of Artificial Limbs