AIFER – Artificial Intelligence for Emergency Response


Artificial intelligence for analysis and fusion of remote sensing and internet data for decision support in emergency management




Disaster events and major damage situations such as floods, forest fires, extreme snow conditions or storms are increasingly presenting disaster control with major challenges in terms of (1) the availability and use of near-real-time and large-scale information for recording and assessing the situation, (2) evaluating the data in near real-time and (3) Fusion of derived information levels for intuitive, transparent and focused decision support. The aim of the AIFER project is to analyze and merge information from innovative data sources (earth observation data and internet data) with the help of AI algorithms, which helps to ensure the protection and rescue of people and the protection of critical infrastructure.
The project approach described aims at the transparent AI-supported and automated analysis and fusion of data from geo-social media, geo-referenced news, Google Trends and earth observation systems, so that end users receive holistic, spatiotemporal information on the situation of the disaster (with transparent information on the strengths of the various sources) is available and the learning algorithms can be optimized by users using a web application via feedback mechanisms.
For this, a comprehensive consideration of ethical, legal and sociological aspects is of central importance for the project: 1.) Research into the legal, ethical and sociological framework of the project developments as well as guaranteeing the acceptance and acceptability of the developed components, 2.) The creation of “Privacy by Design” Guidelines for the proper storage, analysis and dissemination of data from geo-social media, satellites and news data, as well as the analysis results.
AIFER primarily addresses the flood disaster scenario with the analysis of two historical (“cold case”) and one current real-time use case (“warm case”), whereby the transferability to other disaster scenarios is shown on the basis of forest fires, storms and extreme snowfall events becomes.

Key Publications
Kounadi, O., Resch, B., Petutschnig, A. (2018) Privacy Threats and Protection Recommendations for the Use of Geosocial Network Data in Research Social Sciences 7(10), 191, DOI:
Bernd Resch (project lead)
Andreas Petutschnig, Clemens Havas

Project Partners