Energy management for robotic combat vehicles via physics-guided digital twin and edge AI (EXERGI PREDICTIVE) — a federal contract (USAspending)

A robotic combat vehicle is a ground combat vehicle that operates without a person on board. Because such vehicles run on limited energy such as batteries or fuel, how much energy is directed where and when, that is, energy management, shapes how long and how far they can operate. This contract aims to advance that management with two technologies. One is a physics-guided digital twin, which reproduces the vehicle's behavior in a virtual space in line with physical laws so it can be predicted and tested before the real machine moves. The other is edge AI, meaning AI that runs directly on the vehicle's own hardware rather than sending computation to a distant cloud. The aim of pairing them, as the text suggests, is to allow autonomous decisions even in field conditions where communication may be unreliable.

This matters because energy is a constraint that directly bounds an uncrewed vehicle's capability. For the same vehicle, how energy is allocated changes how long and how far it can move, which in turn sets the range of possible operations. If a virtual replica can estimate the best allocation in advance and field-side AI can adjust it in real time, limited resources may be used longer and with less waste. Figures such as how much operating time might be extended are not in the contract text, so they are not addressed here.

Viewed broadly, this research carries implications beyond defense. Combining advance optimization through a digital twin with edge AI that runs entirely on the device connects to any field where the question is how to operate long and smartly on limited power, such as electric vehicles, drones, and industrial robots. Tracking, one contract at a time, how federal research funding is directed toward such foundational technologies offers a way to read both technology trends and the direction of public investment.