Event-Driven Sensing and Embedded Neuromorphic Platforms for Gamma Radiation Monitoring

This work will present an embedded neuromorphic platform developed for long-term unintended gamma radiation monitoring. We describe the hardware architecture and supporting software developed to demonstrate neuromorphic computing for applications where ultra-low power and always-on sensing are required. This is followed by a discussion of our current work on an improved platform that integrates both event-driven vision and gamma-ray spectroscopy sensors for nuclear safeguards applications. Finally, future research directions toward event-driven sampling techniques to integrate analog-to-information reduction into the sensing electronics are proposed.