IdentifAI Tomorrow´s Disasters

Reference number
Coordinator GLOBHE DRONES AB
Funding from Vinnova SEK 1 937 000
Project duration May 2018 - June 2019
Status Completed
Venture Framtidens drönare
Call Drönare för individer och samhälle

Purpose and goal

The IdentifAI Tomorrow ´s Disasters project has developed a cloud-platform where image-data from drones automatically gets analysed by artificial intelligence, through visual image recognition. The project has focused on addressing flooding, the most common natural disaster in the world which causes great damage and cost to local populations, the environment and the economy. Emergency response teams are now able to identify risk areas earlier as well as provide quick, valuable insights once disaster has struck.

Expected results and effects

Disaster relief efforts can now get access to improved site information in natural disaster areas without requiring human analysis to the same extent as before which enables for natural disasters to be dealt with more efficiently and in a safer way. The project result has been tested and verified in the UNICEF humanitarian drone corridor in Malawi and is seen as a valuable tool to get access to automatically generated insights in natural disaster areas in a faster, safer and more efficient way than before. The project result is now being developed further with the goal to scale globally.

Planned approach and implementation

The project has collected large amount of image based drone data and applied different types of CV and ML algoritms to the data. The algoritms have been tested, evaluated, developed and trained to automatically identify, through AI visual image recognition, different objects (roads, houses and water) to more efficiently guide emergency response teams in natural disaster areas. The project result is a cloud-platform enabling for automatic AI analysis of image-data from drones in natural disaster prevention and response.

External links

The project description has been provided by the project members themselves and the text has not been looked at by our editors.