A screenshot of the prototype Caremates developed. There is a dashboard with patients and relevant categories listed.

A project that saves lives! Our DPS team GeoHilfe was presented with an high-impact challenge by our partner DRK Rettungsdienst Bodensee-Oberschwaben GmbH:

Wouldn't it be great if ambulances could head to an emergency right away without any delays?

The interdisciplinary team dove right into research and found out, that
‍
every day, one third of emergency cases in Germany can't be located properly.

This delay can decide over life and death. If an emergency caller does not have access to appropriate mobile services such as GPS or is unable to determine the GPS coordinates due to the extreme situation in which he finds himself, narrowing down the exact location is a process which currently costs valuable time.

The solution: With the help of Artificial Intelligence, GeoHilfe automatically narrows down the possible locations during the emergency call.

GeoHilfe's voice recognition model and speech-to-text transcription analyze the emergency call between the agent in the control center and the patient/caller. It automatically recognizes and extracts location-related keywords.

With the help of a similarity model, the tool shows the caller's most probable location on a map in real time. Intelligent automated suggestions help the agents ask the right further questions to make the localization more precise. After a short and true-to-live conversation on a non-technical level, the agent knows the location of an emergency.

Embedded in a simple user interface, Artificial Intelligence supports the agents to accelerate ambulance dispatchment and save lives.

Dylan, the PM from team Caremates, presenting results at the Final Product Review.

“The Golden Time to make a rescue is typically within 8 minutes, and to describe the proper location of people is the most important task that assures them the basic support of life provided by the rescue team.”

Carsten Schlitz

Quality Manager, Deutsches Rotes Kreuz