The R&D activities address image and laser based identification and verification systems for containment and surveillance; video review systems; and augmented and mixed reality systems. The activities encompass the entire development life cycle from underpinning research and implementation to field deployment. Selected projects are briefly described hereafter:
3D Laser Scanning
Laser scanners generate large amounts of highly accurate 3D data and require advanced software tools for specific application cases. JRC’s in-house processing software enables the inspector to efficiently create as-built maps of nuclear facilities that are used for verifying the design information provided by the facility operator and to monitor changes over time.
Left: snapshot of a 3D model of a (non-nuclear) facility; Right: change map generated by comparing the 3D model acquired before and after modifying the scene. Blue pixels correspond to unchanged objects; red pixels correspond to objects that moved closer to the scan position or were inserted to the scene; green pixels correspond to objects that were moved away from the scan position or were removed.
JRC also developed a Mobile Laser Scanning Platform (MLSP) which is based on Simultaneous Localization and Mapping (SLAM). Additionally to efficient 3D mapping, it provides location information inside nuclear facilities and allows nuclear inspectors to associate all measurements and observations made during an inspection with the corresponding location inside the nuclear facility and thus facilitate subsequent analysis and future inspections. The mapping and localization capabilities form the basis for Augmented and Virtual Reality applications.
Left: Mobile Laser Scanner mounted on a backpack. Right: snapshot of 3D model acquired with the backpack.
Real-time 3D sensors continuously acquire depth information and can be used to complement standard video surveillance systems. JRC’s software analyses the data for relevant events, e.g. changes and movements in a specific area and tracks the objects of interest (e.g. nuclear material containers). Since the analysis software works on measurements in 3D space, event detection is much more robust than optical video surveillance (which is influenced by ambient light conditions) and can be restricted to a pre-defined area of interest (intrusion zone). Multiple 3D sensors can be deployed to cover large areas and they can be integrated with other sensor types (e.g. optical cameras or radiation detectors) to provide an integrated verification approach.
3D Identification and Containment
Using 3D laser triangulation, JRC develops systems for the identification and containment of nuclear material containers. Laser triangulation can measure the surface geometry of an object with an accuracy of several micrometers and therefore allows acquiring a unique fingerprint of the containers. For example, the Laser Mapping for Containment Verification (LMCV) system was developed to verify Dry Storage Casks for spent nuclear fuel. It maps the surface geometry of the weld connecting the lid and the main container body to uniquely identify the container and verify the integrity of the weld.
LMCV is being mounted on a dry storage cask for acquiring a surface scan of the weld.