Frequency and intensity of natural disasters are steadily increasing. Authorities and organisations with security tasks (BOS) are facing increasingly greater challenges with regard to the availability and use of near-real-time and large-scale information for assessing the disaster situation. In addition, there are currently no solutions for automated analysis and fusion of heterogeneous big data, such as Earth Observation and internet data, in such a way that a holistic and dynamically updated situation picture is created.
AIFER develops artificial intelligence (AI) methods to automatically extract and fuse information from satellite, aerial image and drone data as well as from geo-social media and news.
The AIFER project aims to provide targeted, dynamic decision support to the BOS. Information from Earth Observation and internet data is analysed and fused by using AI algorithms in order to support the protection and rescue of people as well as the protection of critical infrastructures. End users such as the German Federal Agency for Technical Relief (THW) and the Bavarian Red Cross (BRK) formulate their requirements, discuss them with developers to then integrate them directly into technical research. Comprehensive consideration of ethical, legal and sociological aspects is highly relevant to the project. Validation and integrability into existing operational processes are tested in a practical way together with end users. AIFER primarily addresses the disaster scenario of flooding, whereby the transferability to other disaster scenarios is demonstrated.
The German-Austrian joint project, funded by the BMBF and FFG, started in February 2021. The German project part is coordinated by DLR.