The US military is preparing to deploy new AI products that promise to automatically identify and track potential targets on the battlefield. But humans continue to make decisions about life and death.
San Francisco AI startup TurbineOne announced Thursday that the Army has awarded its flagship frontline recognition system (FPS) a five-year contract worth up to $98.9 million, ranging from small pilot projects to what corporate spokesman described. Register As a “production scale” contract.
Turbineone describes FPS as a model-independent machine learning platform that automates the evaluation of military targets. The platform ingests data from multiple sensors, such as air, land, sea, space-based images, and signals, to detect battlefield threats such as drones, enemy locations, and other targets.
FPS does it all in a no-code hardware-independent environment that allows the average soldier on the field to “build, retrain and deploy custom machine learning models at the edge without coding.” Most importantly, FPS is designed to work without connecting to the Internet or cloud services. Everything runs on local devices, and soldiers can even provide actionable intelligence to their smartphones and other mobile devices.
US government contract data suggests that TurbineOne has been testing FPS in various military sectors since opening the door in 2021, but not all of its applications are purely on the battlefield. It is also designed to carry out spot debris and other foreign objects at military airfields to prevent accidents.
“Despite the advancements, the current military threat identification process remains highly resource-intensive, costly and difficult to scale,” Turbineone announced its $36 million Series B funding round in a May blog post. Of course, AI is the solution.
“Turbineone offers a no-code, easy-to-use toolkit, allowing fighters to quickly deploy powerful AI tools without waiting for specialized technical support,” Turbineone said. The company said the FPS is already in the hands of Army units in multiple theatres and hopes that the deal announced Thursday will speed up the deployment of the entire military.
Take AI to the battlefield and trust it to identify targets – humans often struggle in combat situations – raise many questions about safety and responsibility, but Turbineone had an answer right away when asked about those specific risks.
According to a company spokesperson, FPS does not suggest whether operators will use data from multiple sensors available to identify and track targets. The company added that the reinvigorating nature of the field can help FPS adapt to changing circumstances without assuming soldiers are always right. TurbineOne said it complies with DOD's responsible AI principles.
TurbineOne also says it has one of the fastest records for small businesses moving from the pilot phase to full-scale DOD contracts. Also, a rapid adoption timeline can see that the Pentagon is still in its early stages and falling to itself to adopt private sector technology, facing a list of unexpected outcomes that AI companies have yet to address. ®
