Information (INF)

Program Area

Within the area of information (INF), the focus lies on autonomous composition of ICT services from whatever resources still available. Especially the role of mobile edge computing as well as distributed and decentralized services to support disaster response and recovery is analyzed. This includes 3D modelling from heterogeneous data sources, detection of relevant events from data stream analysis, etc.

Research Question

How to obtain the best information out of the vast amount of data of the digital city in a timely and situated manner?

Team

Projects

Resilience-Centred Multimodal Data Analysis on Multiple Time Scales (INF1)

This project addresses the processing of information, which is initially not compatible for computer analysis, such as written and spoken language, visual information etc. The project focuses on two aspects: (1) cross-modality, i.e. integrated processing of different forms of information and sources, and (2) cross-timescale integration, i.e. from emergency response in real-time to long-term data (data at rest) that can support preparation for future crises.

Principal Investigators: Iryna Gurevych Stefan Roth

Interaction Concepts for Data-Driven Resilience Planning (INF2)

Information about cities is often complex and abstract. 'Understanding' the data requires not only the adequate analysis and processing by the computer, but also the efficient and user-friendly interaction of humans with this data, especially when critical decisions and actions are based on evaluations of crisis situations. The project aims to build on experience with the display and manipulation of information on a screen (city map) and develop ways to display information in the city itself, via Virtual and Augmented Reality as well as interacting 3D models.

Principal Investigators: Jan Gugenheimer Max Mühlhäuser Christian Reuter

Software Infrastructures for Resilient Digital Cities (INF3)

Existing programming concepts and tools for developing and implementing resilient ICT systems vary substantially for different timescales: reactive or event-driven concepts dominate in the real-time domain, while classical-imperative (process-oriented) as well as functional and declarative approaches are applied in the long-term range (“data at rest”). This project aims to create a new basis for software systems that integrate real-time and long-term, including big data concepts such as probabilistic programming. A data-centric middleware for programming digital cities will be designed. The resilience of the resulting heterogeneous ICT-system will be evaluated by means of simulations.

Principal Investigators: Mira Mezini Zsolt István Thorsten Papenbrock Bernhard Seeger