Accessible to U.S. Military, Local Police
Deep Analytics LLC, a leading developer of artificial intelligence (AI) solutions for the DoD, is building a first-of-its-kind improvised explosive device (IED) image data hub accessible online by military Explosive Ordnance Disposal (EOD) teams, police bomb squads, and other government stakeholders. The EOD Data Hub will enable explosives experts to share photographs and details of devices they find in the field and facilitate development of future counter-IED AI capabilities.
“The EOD Data Hub will enable ordnance disposal experts to do their jobs more safely and effectively,” said Greg Hewitt, Deep Analytics Co-Founder. “The Data Hub will be the focal point of an online community where EOD personnel educate themselves and each other on the many types of explosives they may encounter.”
The DoD Irregular Warfare Technical Support Directorate (IWTSD) in Washington D.C. awarded Deep Analytics the contract to build and maintain the Community Explosive Ordnance Disposal Data Hub.
The challenge of IED identification is that there are so few photographic examples of what the devices looked like before they are detonated. The objective of the Data Hub is to gather and organize photos of intact IEDs just as they appear in the field. EOD personnel can upload photos and annotate device components through the website.
IEDs come in many shapes and sizes and are usually covered with dirt, hidden in trash heaps, or disguised as ordinary objects. They can have visible components in common, including power sources, detonators, and wire bundles. The EOD Data Hub seeks to build a massive searchable photographic archive of as many IED designs and concealment methods as possible.
Deep Analytics is building an online community that will encourage ordnance disposal teams to upload their photos and engage with colleagues. The community will also participate in game-like challenges and competitions that will sharpen their ordnance identification skills.
In addition, the EOD Data Hub will serve as an unclassified database of labeled IED photographs that can be used as training data sets to develop Machine Learning (ML) object detection algorithms. These automated algorithms will be applied to video feeds streaming from stationary or mobile cameras, including drones, to find hidden IEDs along the routes of convoys or at crowded public events before they harm members of the military or the general public.
Deep Analytics, which is at the forefront of introducing ML-based algorithm technology to the EOD community under contract to DoD, is organizing the Data Hub with metadata tags so that uploaded photographs can be queried by type or component. IWTSD is making the database available to approved contractors and other government organizations engaged in the development of IED detection solutions for the military and civilian sectors.
Currently in beta testing, the EOD Data Hub is expected to be fully operational by early 2022.
For more such updates and perspectives around Digital Innovation, IoT, Data Infrastructure, AI & Cybersecurity, go to AI-Techpark.com.
