Unlike fixed-point inspections, mobile inspections lack the structure and process imposed by physical choke points, creating significant variability in the inspection process. Working with the Department of Homeland Security through the Counter Weapons of Mass Destruction Office, Heron Systems developed CURIE to deliver a solution capable of reliable and repeatable data collection, adaptability to scenario and/or environmental changes, and enable remote monitoring and data analytics. Leveraging unmanned ground vehicles (UGVs) as scanning assets rapidly advances state of the art in anomaly detection and aid in closing point-of-entry weaknesses.
Enabling a detailed autonomous inspection process that can self-optimize relieves the burden on the operator from much of the detailed, expert-level knowledge requirements of interpreting sensor outputs, sensor configuration, and subsequent actions. CURIE creates a complete synergistic ecosystem with both hardware and software to conduct CBRN inspections implementing the best practices from academia and industry.
Robust workhorse platform capable of performing the broader inspection mission. See vehicle diagram
Mobile-capable user interface empowers inspection agents to confidently conduct the screening process in a supervisory capacity; allowing for intervention where human intuition and perspective drive a need. See prototype screens
Smart displays continuously keep operators informed but not overburdened; providing contextual awareness, data analytics, and clear indication of the robot’s planned actions and workflow steps.
Localized data registration techniques provide continuity between subsequent screening operations and situational awareness for inspection personnel.
Closed-loop feedback mechanisms between the CBRN sensors and navigation systems create rigid and deterministically defined behaviors through informed search strategies; generating granular inspection details with significantly high statistical confidence in the most abbreviated manner possible.
Cloud synchronized operational data and autonomous population of ANSI 42.42 compliant reports for reachback; delivering the necessary meta-data enabling contextual awareness for remote analytics.
Backboned via the MACE robotics architecture promotes a mature autonomous infrastructure that can remain platform agnostic (UxV) and significantly scalable.
Mr. Kroeger and Dr. Kevin Kochersberger had laid the foundations for this work during a multi-year DoE/DoD contract, exploring the application of collaborative UAVs and UGVs for the purposes of Aerial Nuclear Materials Detection and Sampling. The collective research program demonstrated significant benefits that autonomy could provide during passive sensing techniques to localize and distinguish multi-point sources. Improvements in anomaly detection would simultaneously exploit contextual environmental information and probabilistic threat models to refine estimates. These techniques have been demonstrated by the Unmanned Systems Lab of Virginia Tech at SRNL, INL, and TEAMS test sites.