This post is also available in: heעברית (Hebrew)

A groundbreaking under vehicle inspection for explosives and early warning system under development will improve security through artificial intelligence technology. 

Synthetik Applied Technologies has been awarded a $1Million grant from the U.S. Department of Homeland Security (DHS) Science and Technology Directorate (S&T) to continue development of its DeepVIEW system, which leverages state-of-the-art deep machine learning (ML) and computer vision (CV). 

The system will be deployable on very low-cost commercial-off-the-shelf equipment, and modern mobile devices.

The company has already demonstrated the potential of the technology to DHS, and this grant will now allow it to deliver a step-change in current capability creating a system that will learn, evolve, and improve over its lifetime. 

The system is designed to make use of high-resolution, multi-vantage point imaging to provide real-time anomaly detection, and change detection for both automatic license plate reading (ALPR) and under-vehicle inspection and scanning applications in fixed and mobile configurations, according to the company website.

The award builds directly on ongoing programs with DHS S&T which include the development of artificial intelligence (AI) object detection models, and the generation of synthetic training data for explosive detection machine learning algorithms.

The funding is part of the Vehicle Inspection for Early Warning (VIEW) initiative, designed to develop a low-cost, automated, under-vehicle inspection system for roadway mounting at facility access points, road border crossing locations, checkpoints and other vehicle inspection locations.

The company is currently leading research with the U.S. Defense Advanced Research Projects Agency (DARPA) to fundamentally redefine how we process 3-D data using artificial intelligence.  

They are also working directly with NOAA and Microsoft AI for Earth to develop a low-cost entanglement mitigation system to protect endangered marine species.