The US Air Force has asked Colorado-based Stratom to develop a machine-learning method for tracking military assets.

The company is developing an autonomous asset-tracking system (AATS) using artificial intelligence and machine learning for camera systems to identify and track military logistics assets accurately across various environments and modes of transport.

“Our innovative software stack leverages the latest advances in machine learning to detect and transcribe airplane tail numbers and determine which plane the tail number is associated with,” said Elizabeth Gilmour, senior robotics perception engineer at Stratom. “Next, we’ll be able to communicate that information to databases and AGVs to autonomously refuel or load the aircraft. By eliminating the complexities and uncertainties associated with traditional asset-tracking systems, our streamlined, modification-free option simplifies distributed operations while addressing imminent logistics challenges created by adopting the agile combat employment concept.”

The initial phase of the research project focuses on the software component, laying the groundwork for a suite of machine-learning models that will process images to extract data such as asset type and identification numbers. With this approach, the developer of autonomous ground vehicles and robotic systems for commercial and military applications eliminates the need for physical modifications to current assets, offering cost-effective and scalable logistics operations. In addition to automated asset tracking, the technology simplifies the orchestration of logistics equipment for cargo, munitions and refuelling on the flight line.

Initially envisioned as a system to help autonomous refuelling and cargo-loading vehicles identify the correct aircraft, AATS has received interest from several groups for other potential applications. The text spotting could be used to identify not only airplanes but any asset with an identifying number, such as rail cars, trailers or pallets. Further, other users have shown interest in this capability to identify relevant text in a wide range of operating situations.

“Our vision with the AATS extends beyond current logistics,” said Mark Gordon, CEO of Stratom. “Building on our existing machine-learning capabilities in object detection and text recognition, we’re solving today’s problems and anticipating the future needs of a rapidly evolving global market. This sets the stage for our machine-learning-powered system to seamlessly integrate into various sectors across military and commercial applications, providing unprecedented tracking accuracy and operational efficiency.”

Stratom (www.stratom.com) develops unmanned ground vehicles and autonomous robotic systems for commercial and military applications. Specialising in unmanned cargo movement, robotic refuelling, robotic hazardous liquid transfer and autonomous mobile robots, the company’s military-proven tools, methods, technologies and services can solve difficult logistics and operational problems.

The Air Force Research Laboratory (afresearchlab.com) is the US Air Force’s primary scientific research and development centre. With a workforce of more than 12,500 across nine technology areas and 40 other operations across the globe, it provides science and technology ranging from fundamental to advanced research and technology development.