Automatic situation detection to increase the civil safety of people in hazardous situations

Application industry: Public sector
Technology area: Machine learning

SmartSense&Rescue is developing a smart system that automatically locates people in life-threatening situations using video images from camera drones. The system supports emergency services on-site and thereby increases the safety of citizens in the event of disasters.

The baseline
In dangerous situations, such as natural disasters, terrorist attacks, or mass panics, emergency services must respond quickly and effectively to save human lives. Much depends on whether the situation can be clarified precisely: What are the conditions on-site? How many people are in danger? Where are these people located and where are they moving to? Drones and camera systems are already frequently used in such cases. However, this technology has so far only been used for purely visual live observations, as an ‘extended eye’ as one might say. This is because the systems feature almost no smart functions due to the lack of software-based evaluation options. At the same time, the hazard situation in Germany is worsening – for example, due to the constantly growing threat of forest fires – and first responders are increasingly reaching their limits

The project goal
SmartSense&Rescue is developing a smart software solution to automatically count, localize, and trace people’s movements based on video recordings from commercially available drones. Existing algorithms are now being further developed and adapted, integrated into a complete system, and optimized for rescue operations. All the information delivered from video cameras, thermal imaging cameras, or GPS transmitters, for example, is evaluated and visually processed almost in real time. Emergency services can use colour codes in a 3D map to determine the level of danger a person is in, at their current location. The solution is to be integrated into existing operational command systems of authorities and organizations performing security tasks and thereby support emergency forces on-site. SmartSense&Rescue will be tested and evaluated in several scenarios together with the application partner, the fire department of the district of Cham.

Application and practical benefit
The smart technology is intended to reduce the workload for first responders and is also more robust and accurate than existing individual procedures. SmartSense&Rescue can be used in a wide range of scenarios, from searches for missing persons to severe disasters and complex dangerous situations, such as floods or industrial fires.

Term: 1 May 202 to 01 April 2022

CrisCom Solutions GmbH (lead member), Technische Universität Berlin, Cham district authorities

Klaus Lino Ruatti
CrisCom Solutions GmbH