The proposed work will advance the themes of Cognitive and Neural Processes in Realistic, Complex Environments and Data-Intensive Neuroscience and Cognitive Science. We propose a high-risk, yet rigorous and principled approach towards understanding neural processes in realistic, complex environments. Specifically, we will develop a set of bio-realistic hardware platforms to systematically explore the neural basis for sensorimotor control in the vibrissal (whisker) system. Hardware offers a method to gather and synthesize vast amounts of data that cannot be obtained through the study of animal behavior, neurophysiology, or through simulation; it is an approach to data-intensive neuroscience that exploits true parallelism to model neural and sensorimotor processes across complexities of scale and environment. Hardware models expose how biological individuality and variation ensure behavioral robustness, because they inherently incorporate trial-to-trial motor and sensing variability, and can reveal multiple families of control circuits that could underlie a given sensorimotor strategy. . Integrative Value and Transformative Potential: The proposed work will help bridge the gap between 'sensory' and 'motor' systems that has characterized much of neuroscience for the past 40 years. Our high-risk, high-reward plan to build the sensorimotor system of an animal for the specific purpose of understanding neural computation relies on sensing at scales relevant to real animal and the development of new technology that allow us to use the animal's own sensors on the robot. Our work will lead to transformational insights into the neural basis by which 'low level' but powerful neural circuits confer animals with flexibility and resourcefulness in multiscale sensing and actuation.
|Effective start/end date||9/1/17 → 8/31/21|
- National Science Foundation (BCS-1734981)