Information about the condition of battlefield casualties could be improved by ATRACT, a project led by Edge Hill University to develop an AI-enabled autonomous triage drone.

Edge Hill University

Funded by EPSRC, ATRACT (A Trustworthy Robotic Autonomous system to support Casualty Triage) will see the development of a flying drone that can assist and speed up battlefield triage. The £850,000 project concludes in 2026.

Ardhendu Behera, Professor of Computer Vision and Artificial Intelligence (AI) at Edge Hill, explained that frontline army medics are often required to monitor multiple casualties and prioritise them based on the severity of injuries. Consequently, there is an urgent unmet need for enhancing casualty survival in a warzone where conventional helicopter evacuations are not possible, due largely to an abundance of cheap, accurate shoulder-launched ground-to-air missiles .

The project will focus on four main objectives that represent major innovations in the use of AI and RAS (robotics autonomous systems).

The first is to develop advanced sensors so that ATRACT can accurately search for injured soldiers using visual and thermal imaging data while still manoeuvring over and around challenging terrain.

The second and third objectives are focussed on the data ATRACT collects. The research team will combine advanced multimodal AI sensing and advanced algorithms to detect the location of frontline soldiers and provide real-time monitoring of soldiers' injury severity and their vital signs for effective triage management.


ATRACT will then have to provide real-time casualty information to a medical team as it approaches, enabling more effective resource management and casualty prioritisation.

“Even though its autonomous, we’ll try it out a simulation environment before we send it out,” said Prof Behera. “We’re trying to solve some of the basic requirements, for instance the dynamic nature of the scene will change continuously.

“Our main target is for the drones to be close enough to the soldiers to capture all the vital signs to provide information for triage.”

He added that ATRACT initially aims to prove the technology on a single drone before extending the concept to a team of drones capable of working collaboratively.

IP that could be generated by the project involves being able to locate soldiers in a camouflaged scene, using enhanced computer vision and AI algorithms to locate them.

Loughborough University, and the Universities of Brighton and Portsmouth are supporting ATRACT.