RESCUER: "Crowds should take an active part in ensuring their own safety during major events"

Interview with Dr. Karina Villela, Fraunhofer Institute for Experimental Software Engineering IESE, and Prof. Paul Lukowicz, German Research Center for Artificial Intelligence

Thousands of people push through a tight tunnel: 21 people died while several hundred people were injured this way during the Love Parade 2010 disaster in Duisburg, Germany. Today we know that such disasters could be prevented, if communication between event participants and rescue workers would be better.


Photo: Smiling woman with dark hair and striped blouse - Dr. Karina Villela

Dr. Karina Villela; ©Fraunhofer IESE

This is why now a communications platform is being developed during the EU-Brazil RESCUER project, with which these types of emergencies are supposed to be prevented during major events. The project participants are working on an intelligent information technology solution with which event participants themselves are able to contribute to their own safety. spoke with Dr. Karina Villela from the Fraunhofer Institute for Experimental Software Engineering IESE and Prof. Paul Lukowicz from the German Research Center for Artificial Intelligence (German: Deutsches Forschungszentrum für Künstliche Intelligenz, DFKI) about the communication solution and its implementation during the 2014 FIFA World Cup in Brazil.

Dr. Villela, with the international RESCUER project you are working on a communications platform that is meant to save lives. How can we picture this?

Dr. Karina Villela: Incidents occurring during large-scale events and in industrial areas may have a huge impact on human lives, property, and the environment. In critical situations, fast reaction is vital in order to avoid any damages. Just think of the incident at the Love Parade in Duisburg in 2010. The main challenge for a command center is to quickly obtain contextual information about an emergency situation in order to make the right decisions.

The project RESCUER intends to allow eyewitnesses, first responders, as well as formal responders to provide information to the command center at the place of the incident in a way that is appropriate to their own exposure to risks. In such a critical situation, thousands of images, videos, or texts might be sent to the command center. Therefore, RESCUER will also have the capability of semi-automatically filtering, merging, and analyzing multimedia data so that only relevant information is provided to the command center accelerating the decision-making process.
Photo: Smiling older man with grey suite - Prof. Paul Lukowicz

Prof. Paul Lukowicz; ©DFKI

Furthermore, RESCUER will support the command center in providing official and accurate announcements of emergencies to the affected community and to the general public. Finally, the ad-hoc communication between people in the crowd will be supported in the case that the communication infrastructure is overloaded or breaks down. RESCUER will be much more than a communication platform.

Prof. Lukowicz, what role do the Fraunhofer IESE and the DFKI play in the RESCUER project?

Prof. Paul Lukowicz: The Fraunhofer IESE is the project manager for the consortium of nine German and international partners. The team’s task is to analyze human behavior in emergencies in industrial areas and during major events and to develop an interaction strategy and corresponding user interfaces to support the safe and efficient information transmission through the crowd at the event location. The Fraunhofer IESE is also going to create a data quality model to filter and consolidate information on the crowd, so other partners can subsequently develop efficient data analysis methods.

My team at the DFKI in turn focuses on the technology that enables software use with a mobile radio unit and communication between control center and apps. These crowdsensing and management systems were used globally during major events such as the 2012 Olympic Games in London or the coronation of the Dutch King in 2013 in Amsterdam. We are also developing a so-called ad hoc network that permits data exchange within a crowd, even when the official network is not available.

The communications platform is particularly designed for crisis and emergency management use during major events like the 2014 FIFA World Cup in Brazil. Why is it so well suited for this?

Villela: Because there is a crowd attending a large-scale event and they normally take along their mobile devices. They outnumber the security officers, firefighters, and rescue service officers at alarge event by far, and they are spread all over the event venue. In this sense, the command center can have a pair of eyes everywhere. Those people are able to report an incident/problem at its very beginning. Thus the command center and the involved operational forces can deal with the incident/problem much earlier, before it has become a critical situation.
Photo: Spectator in the stadium takes a picture with his mobile phone; Copyright: Tyler Olson

The new communications platform gives people the opportunity to have an influence on their own safety; © Tyler Olson

What technology is behind the RESCUER solution?

Villela: Several techniques will be combined. Regarding the communication infrastructure, DFKI and our Brazilian partner Universidade Federal da Bahia (UFBA) will develop a peer-to-peer communication method based on the built-in WiFi capability of mobile phones. The idea is that a certain number of smartphones would turn into WiFi access points, creating local networks.

The Fraunhofer IESE will define a variation management strategy for the mobile applications the RESCUER platform consists of in order to not just consider the different Smartphone models used by people in the crowd, but also other variations like for instance the special scenario (major event or industrial area) and the role of the person in the crowd (eye witness or rescue worker).

Other project partners like the University of Madrid and the University of São Paulo will investigate how current image and video analysis techniques can be extended or new techniques can be developed to deal with the challenge of multimedia data analysis in an emergency situation.

How user-friendly is the new communications platform?

Lukowicz: The usability of the RESCUER platform is a key factor for its success, especially during a disaster. This is precisely why the Fraunhofer IESE defined a three-part strategy for gathering data in a crowd under stress: 1. fully automated sensor data collection, 2. partially interactive data collection, and 3. fully interactive data collection. How and what data is being gathered depends on how much stress there is at the time. The Fraunhofer IESE will combine the latest technologies for the development of mobile applications as well as user experience to help people under stress in using their Smartphones.
Photo: Tightly packed stadium ranks; Copyright: Ludwig

The photos, videos and text messages delivered by the crowd will be semi-
automatically filtered and forwarded to the command center; © Marion Ludwig

What happens to the collected data after the event?

Lukowicz: The data we collect with the help of sensors is not individual, but rather aggregate data pertaining to crowd density or crowd flow. The analysis of this data can provide valuable clues for handling future disasters. The data that people send on their own to the command center in the form of pictures or messages are always subject to data protection regulation. The permission to store and use the data has been obtained in advance.

What significance does the term crowdsourcing have in this project?

Villela: In the strict sense, crowdsourcing means "coming from the crowd voluntarily". But in a broader sense, crowdsourcing information in RESCUER means empowering citizens to actively participate in ensuring their own safety and security.
Foto: Michalina Chrzanowska; Copyright: B. Frommann

© B. Frommann

The interview was conducted by Michalina Chrzanowska and translated by Elena O'Meara.