ABSTRACT
BRIDGE is a European collaborative project established within the Security Research sector of the European Commission. The basic goal of BRIDGE is to contribute to the safety of citizens by developing technical and organisational solutions that improve crisis and emergency management in EU member states. A (middleware) platform is being developed that is to provide technical support for multi-agency collaboration in large-scale emergency relief efforts. Several tools and software systems are being implemented and tested to support first responders in their efforts. Beyond technical considerations, organisational measures are being explored to ensure interoperability and cooperation among involved parties; social, ethical and legal issues are being investigated as well. A focus of the project is to demonstrate and validate its results in the course of real-world emergency response exercises.
Since most of the BRIDGE work is beyond the scope of this e-letter on social networking, only a brief overview of the BRIDGE goals and work will be given. However, one thread of work is relevant in the context of social networking and deserves to be covered more closely: automatic detection of notable sub-events of a crisis from social networks. This activity makes use of crisis-related information coming from citizens via social networks and thus contributes to building an improved operational picture in a crisis situation and to better planning and performing crisis response tasks.


1 INTRODUCTION: BRIDGE OVERVIEW
The BRIDGE project [1], started in April 2011 and planned for a duration of four years, set out to support interoperability (build a 'bridge') between multiple first responder organisations and thus to contribute to effective relief efforts in large-scale emergencies such as natural catastrophes, technological disasters or terrorist attacks. The vision of the project is to:

• facilitate cross-border and cross-agency collaboration;
• allow the creation of a common, comprehensive, and reliable operational picture of the incident site;
• enable integration of resources and technologies into workflow management;
• enable active ad-hoc participation of third parties.
  

The BRIDGE consortium consists of 14 partners, bringing along a spectrum of competencies in:

• technical areas: Stiftelsen SINTEF, NO (project coordinator); Almende B.V., NL; CNet Svenska AB, SE; Fraunhofer Institute for Applied Information Technology FIT, DE (technical coordinator); Thales Nederland B.V., NL; Alpen-Adria-Universität Klagenfurt, AT; Paris-Lodron-Universität Salzburg, AT; Technical University of Delft, NL;
• training/exercises: Crisis Training AS, NO; SAAB Training Systems, SE; VSH Hagerbach Test Gallery LTD, CH;
• social, ethical and legal issues: Lancaster University, UK; Stockholm University, SE; and
• domain/end user requirements/expertise: SINTEF, NO; Fraunhofer FIT, DE; VSH Hagerbach Test Gallery LTD, CH; Helse Stavanger HF, NO.
  

The project work, originally destined toward the individual work packages’ objectives, was directed early on toward collaboration across the work packages and toward achieving the goals of major demonstrations planned for the project. Demonstration 1 (Sept. 2012) showed interoperability techniques and measures for individual components in the course of a fire fighter exercise in the VSH Hagerbach Underground Test Gallery, Flums, Switzerland; Demonstration 2 (April 2013) was a table-top presentation of improved components and technologies, with a special emphasis on visualisation and interaction techniques; Demonstration 3 (Sept. 2013), prepared by Demonstration 2 and specifically showing BRIDGE collaboration tools and support, was held in the course of a real, non-trivial training exercise in Risavika Harbour near Stavanger at the west coast of Norway, involving police, fire fighters and paramedics working together to cope with a (simulated) terrorist attack. The demonstrations are described in some detail in the BRIDGE newsletters available from [1]. The project is supported by an End-User Advisory Board (EUAB) with members from European emergency response and public safety organisations, who provide valuable guidance and feedback on the directions and developments of the project.

Work before and between the demonstrations proceeds along the following lines, in each cycle refining and extending existing tools and results, in some cases also introducing new ones:

• domain analyses and requirements engineering, mainly done by means of co-design workshops among BRIDGE project members, EUAB members and end users/practitioners;
• design and development of technical components;
• integration, validation and evaluation, partially again involving EUAB members and end users/practitioners; and
• demonstrations themselves to end users/practitioners, with the goal to receive feedback and advice on the utility and usability of the tools, systems and procedures and to finally create solutions that will be useful and have impact.
  

In order to foster progress in the project and promote collaboration among work packages, particularly between developers and non-technical people, so-called 'Concept Cases (CCs)' were devised and are being deployed in BRIDGE. A Concept Case (CC) is a specific novel tool, system or procedure that meets certain needs of some participant(s) or stakeholder(s) in crisis response operations, for instance improving the triage process, enhancing the operational picture/situational awareness, easing communication and collaboration of different organisations, better managing resources, or involving the public in the response activities. CCs are the instruments used in BRIDGE to advance the project and to align its results with the users' requirements and tasks.


2 STATUS AND ACHIEVEMENTS: CONCEPT CASES
The status and achievements of the project so far can probably be best explained by means of the current CCs, shown during Demonstration 3 (Sept. 2013). More information can be found in the BRIDGE Newsletter, Issue 4 [2] and in more detailed descriptions of the CCs [3].

The BRIDGE Middleware is a special component destined to take a central role in 'glueing together' BRIDGE (or other) emergency response (sub-)systems, among them those developed as CCs. The middleware is service oriented, offering services in the broad classes: communication, orchestration, data/model management, and security/trust. Services that were already demonstrated include: messaging, publish/subscribe, network management, triggering and eventing, workflow management, service catalogue, identification, shared data store.

The major BRIDGE CCs and brief explanations are as follows:

• Adaptive Logistics: This CC organises dynamic, multi-agency collaboration by means of workflows, specifically a workflow generation and management tool that plans, instantiates, monitors and adjusts human or/and virtual activities.
• Advanced Situation Awareness: This CC aims at improving situational awareness by supplying real-time visual and other information on the extent and nature of the disaster and its consequences. A UAV (hexacopter) provides live video feeds and real-time environmental monitoring data; an expert system analyses the environmental data and supports decision making; and a modeling module delivers estimates/simulation results of structural damages due to explosives and of the dispersions of plumes in case of uncontrolled releases.
• Dynamic Tagging: The dynamic tagging system enables first responders to mark and monitor specific locations of an incident site and to create enhanced situational awareness. Marking of points of interest may be done physically, via sensor tags, or virtually, using for instance symbols, text, or voice recordings on a map. Response personnel may visualize the information items left by other persons on mobile devices using a map-based or an augmented reality view.
• First Responders Integrated Training System: The objective of this CC is to establish a well-structured learning and training methodology, supporting training, evaluation, simulation and quality improvements in crisis management operations.
• Robust and Resilient Communication: When disaster strikes, communication and networking facilities will be at least partially affected and become non-operational. This work thus provides a wireless mesh network that establishes ad-hoc networking services on an incident site, as far as physically possible. For demonstration purposes, a 'help beacon' application for Android smartphones was developed, allowing people to call for help over the ad-hoc network, and a counterpart “help beacon seeker” is available, collecting help beacons in the vicinity and enabling first responders to locate and rescue victims.
• Situation Aware Resource Management: By monitoring and tracking the locations (as well as other context information such as activities) of emergency personnel using a smartphone app, this CC enhances resource management during response operation, for instance dynamic team formation or task assignment to individuals.
• Master: This CC is the central tool presenting/visualising, and allowing to act upon, information about an incident (e.g., location, victims), about the response (e.g., personnel, vehicles), and from external sources (e.g., weather). The operational picture can be presented on a large touch-sensitive table (e.g., for use by the incident command team), on a tablet (e.g., for use by a team leader), or on a PC (e.g., in an operational center).
• Information Intelligence: This CC collects data posted by people in social networks or from the incident site, including Tweets, images and videos tagged with textual annotations, analyses the data, identifies and clusters sub-events (specific hotspots of a crises), and visualizes the results on a map. Since this CC pertains to social network analysis, it will be described in more detail in the next section; technical information is given in the article by Pohl et al. in this e-letter.
  

Further work associated with BRIDGE explores ethical, legal and social consequences of new practices of public engagement in emergency response via social media. A contribution on this by Büscher et al. can be found in this e-letter. In particular, that article discusses the challenges and opportunities of the 'socio-technical innovation' of involving the public in formal crisis management operations via social media.


3 SOCIAL NETWORK ANALYSIS: THE 'INFORMATION INTELLIGENCE' CONCEPT CASE
Many people nowadays document any remarkable situation they experience, in the form of text snippets, images or even videos, via social networks, special sharing platforms or private communication. This behaviour also pertains to disaster situations if the individual circumstances allow. It is thus obvious that this information provided by citizens either in social networks or communicated from an incident site can be used to improve the operational picture, keep it up to date, or even extend it by detecting critical (sub-)events within the overall crisis at an early stage.

The BRIDGE Information Intelligence (II) CC thus aims at developing and testing a tool that automatically analyses social media data and live data from the emergency field, identifies/clusters specific sub-events of the emergency, aggregates the results, and visualizes them on a map to provide a timely situational report. The II tool currently comprises the following components:

• Data Simulation Component: For demonstration purposes (during Demonstration 3 at the Stavanger exercise) and for later use for training, testing and evaluation tasks, this component was developed to create (simulated) incident-related 'social network' datasets, i.e., short messages (e.g., simulated Tweets) that model social network activities (e.g., Twitter) pertaining to an incident. Based on an XML scenario description, pre-defined textual phrases, sub-event attributes, and temporal specifications, this 'social media' traffic can be synthesised and integrated into a running exercise. This component has such received some attention from first responders since this concept appeared promising to them for simulation and training purposes.
• Data Collection Component: A smartphone app was implemented to enable people on the incident site to contribute images and text messages to the dataset and integrate them into the analysis process.
• Data Analysis/Aggregation Component: This central module analyses and aggregates the data based on location, time and textual information using online clustering algorithms. The goal of the analysis is to find relevant sub-events in the crisis and to spot/visualize them on a map and related to a timeline, including the option to get a detailed view of the sub-events, i.e., of keywords and phrases characterising them; Fig. 1 shows an example. The visualisation can be done via any browser or sent to the Master for display to the command personnel in order to update their operational picture. The technical background, datasets and details are described in the article by Pohl et al. in this e-letter.
  

Fig. 1. Visualisation of detected sub-events (markers by Map Icons Collection; mapicons.nicolasmollet.com)


Demonstrations to and interviews with members of the EUAB and personnel from several agencies provide some confidence that information and insights gained by means of social media monitoring are regarded as helpful. A survey conducted among practitioners from several European countries shows the practical applicability of a (sub-)event detection tool like the II CC during emergency management. Results are reported in [4].


4 FUTURE WORK
The BRIDGE project progresses toward an integrated Demonstration 4 near the end of the project (March 2015). The social network analysis tool (II CC) will be extended to deploy and test further online clustering algorithms as well as to consider trustworthiness of the information contributed by the public, e.g., by classifying the messages received or data retrieved before processing them.


Acknowledgment
The research described in this article has received funding from the European Union Seventh Framework Programme (FP7/2007-2013) under grant agreement nr 261817 and was partly performed in the Lakeside Labs research cluster at Alpen-Adria-Universität Klagenfurt. Most of the material in this overview is taken, and publicly available, from the BRIDGE Website [1], mainly in the BRIDGE newsletters, deliverables and publications. The contributions of BRIDGE consortium members (text, figure) are greatly acknowledged.


REFERENCES

[1] BRIDGE Website. http://www.bridgeproject.eu. Accessed: Feb. 2014.

[2] BRIDGE Newsletters. http://www.bridgeproject.eu/en/promo-material/bridge-newsletter. Accessed: Feb. 2014.

[3] BRIDGE Concept Cases. http://www.bridgeproject.eu/en/bridge-results/concept-cases. Accessed: Feb. 2014.

[4] Daniela Pohl, Abdelhamid Bouchachia, Hermann Hellwagner. Supporting Crisis Management via Detection of Sub-Events in Social Networks. In: International Journal of Information Systems for Crisis Response and Management (IJISCRAM), vol. 5, no. 3, pp. 20-36, July-Sept. 2013.



Hermann Hellwagner is a full professor at the Institute of Information Technology, Klagenfurt University, Austria, leading the Multimedia Communications group. His current research areas are distributed multimedia systems, multimedia communications, and information-centric networking. He has published about 200 scientific papers on parallel computer architecture, parallel programming, and multimedia communications and adaptation. He is a senior member of the IEEE, member of the ACM, and Vice President of the Austrian Science Fund (FWF).