Voice-commanding a UGV: Best Paper finalist at SSRR 2025

November 2025 · Presented at IEEE SSRR 2025, Galway

‹ ›

In late 2025 we travelled to Galway, Ireland, to present the first paper from the NAMUR project at the 2025 IEEE International Symposium on Safety, Security, and Rescue Robotics (SSRR): Real-World Deployment of an LLM-Enabled Voice-Commanded UGV for Logistics in SAR Missions. We were honored that the paper was a finalist for the Best Paper Award 🏆.

Talk to the robot

Traditional robotic command interfaces are often rigid and cognitively demanding — exactly the wrong properties under the pressure of a live search-and-rescue mission. The NAMUR project addresses this by leveraging large language models to translate intuitive human commands into coordinated robotic actions, reducing operator workload and improving flexibility.

In this paper we demonstrated how operators can give natural-language voice commands — such as "deliver the batteries to the drone pilot" — that are safely interpreted and executed by uncrewed ground vehicles (UGVs) in real SAR missions.

Validated with real first responders

The system was validated in realistic search and rescue missions with over 150 first responders, showing significant improvements in mission efficiency and operator experience through safe, context-aware autonomous navigation and task execution.

A collaboration between SDU and the University of Málaga (IMECH.UMA), the work reflects a strong international partnership — and the demonstration video was also showcased at the XIX Jornadas Internacionales de la Universidad de Málaga on security, emergencies, and disasters.