Real-time radiation monitoring is critical for public health and emergency response. High-frequency monitoring can generate large amounts of data for dozens of radioactive isotopes though. There is a growing demand for compact radiation detection devices that are also able to quickly and autonomously process these large datasets for anomalies. A UC Santa Cruz researcher has developed machine learning software that synthesizes real-time radiation monitoring data in situ to detect radioactive anomalies.
A UC Santa Cruz researcher has designed software that is used in line with a radiation detector to identify radioactive isotope anomalies. The software uses a field-programmable gate array-based neuromorphic architecture and a spiking neural network to synthesize and display real-time anomalies in radioactive isotope spectra data. This technology is compact, portable, and low-power, and can be used for unmanned and unmanned aerial monitoring.
Public health emergencies
Radiation Monitoring and detection
Compact, portable, low power
Fast processing times
Low detection thresholds and data storage needs
Compact, portable, low powerAutonomous processing
Radiation Detection, Machine Learning, Ambient Monitoring, Nuclear contamination, UAV, Drone