Stay Informed:
Baskin Engineering COVID-19 Information and Resources
Campus Roadmap to Recovery
Zoom Links: Zoom Help | Teaching with Zoom | Zoom Quick Guide

Advancement: Feedback Control in Neuromorphic Systems

Speaker Name: 
Alex Spaeth
Speaker Title: 
PhD Student
Speaker Organization: 
Computer Engineering PhD
Start Time: 
Monday, April 12, 2021 - 3:00pm
End Time: 
Monday, April 12, 2021 - 4:00pm
Location: 
Zoom - https://ucsc.zoom.us/j/99321772161?pwd=WEw1ZWMzbjNsTjMzd21HQ0pOS2ExUT09 - Passcode: 320678

Abstract: So-called "neuromorphic" engineering, where artificial systems are designed based on evidence from biology, has become increasingly popular in recent years, spurred on by successes in areas such as machine learning and brain-inspired neuromorphic hardware. However, typical neuromorphic approaches are focused on the highest levels of cortical function, attempting to recover the advantages of human cognition in a computational setting before gaining full understanding of the complex adaptive behavior available even to much less sophisticated biological neural systems. Research on biological motor control is rarely applied to artificial systems, and even such applications are rarely placed in the context of the existing theory of feedback control.

This advancement proposes a thesis presenting an approach to closed-loop control in neuromorphic systems. Theoretical convergence guarantees will be developed as an extension of the theory of event-based control, and some applications will be explored. Neuromorphic feedback is applied to a designed neural circuit in order to construct a state machine which generates a crawling gait on a modular robot. Additionally, closed-loop neuromorphic position control is combined with spiking reinforcement learning to demonstrate the applicability of these techniques to more demanding physical robotic platforms.

Event Type: 
Advancement/Defense
Advisor: 
Mircea Teodorescu
Graduate Program: 
Computer Engineering PhD