ELROB Puts Ground Robots Through Field Trials

Armasuisse, Switzerland’s federal defence procurement office, hosted the 2026 European Land Robot Trial from June 15 to 19 at the Thun military training area. The event, organised with Germany’s Fraunhofer FKIE and RUAG, is not framed as a trade show or a race for spectacular demos. It is a field trial for unmanned ground vehicles. More than twenty international teams are testing robotic systems in realistic scenarios, with the stated aim of assessing what ground robotics can already do in demanding civil and military operations.

The structure is the useful part. ELROB splits the trial into five use cases. Robots have to support convoy transport, reconnoitre non-urban terrain, follow a soldier as a “mule” over roughly 300 metres, assist in search and rescue or casualty evacuation, and identify explosive remnants of war. In the convoying scenario, at least two vehicles must reach a destination, each vehicle should be able to carry a payload of at least 1,500 kilograms, and the task should be completed with as much autonomy as possible. Those constraints offer a clearer signal than isolated robot videos because they stress navigation, payload, perception, endurance and human integration at the same time.

That is why the trial matters beyond defence procurement. Ground robotics receives less public attention than aerial drones or humanoids, but its near-term value is practical: scouting dangerous terrain, moving supplies, evacuating an injured person, or inspecting a threat before a human approaches. Armasuisse emphasises practical testing because moving from lab conditions to field use requires machines that can deal with dust, uneven ground, obstacles, unreliable links and operators under time pressure. The promise is not a fully autonomous wonder machine. It is a tool that can reduce human exposure in risky tasks.

ELROB also shows how civil and military robotics are beginning to share the same building blocks. Two or three-dimensional mapping, person following, semi-autonomous transport and hazardous object detection can be relevant after a disaster as well as in a conflict zone. That overlap calls for a sober reading of progress. The value of a system is not just the sophistication of its onboard intelligence. It is whether the machine can perform a risky task repeatedly, be assessed by a jury, and fit into existing procedures. For European robotics, this kind of trial says less “look at the future” than “measure what already survives the field”.