The collection and analysis of reliable data captured by monitoring equipment located at nuclear facilities is essential for international nuclear safeguards. However, the current safeguards systems have expensive data encryption and authentication techniques that can restrict broader application and the ease of use of these technologies. This research aims to enhance trust in the anomaly detection capabilities within the surveillances and exposure of breach data. Our approach leverages two technologies: Distributed Ledger Technology (DLT) and grammar compression (GC) based anomaly detection. DLT creates an immutable and decentralized system to monitor the origin and flow of data within safeguard enclosures providing full transparency. GC is used to monitor multivariate time-series data gathered at the field through sophisticated measurement devices and identify breaches with minimal resource utilization. Also, investigation on the application of multi-party computation (MPC), which allows the secure integration of other operator data, such as safety or physical protection system logs, without revealing confidential or proprietary data, is conducted. The methods were evaluated using the data from the Multiple-Informatics for Nuclear Operations Scenarios (MINOS) testbed, which is a realistic proxy for international safeguards data. Preliminary findings support that this integrated framework has the potential to improve the efficacy, security, and scalability of verification processes, as well as enabling the integration of novel data types while upholding privacy concerns.
 

Anomaly Detection, Distributed Ledger Technology, Multi-Party Computation, Privacy-Preserving Data