Abstract: The latest EU funding framework, ?Horizon 2020?, has moved consideration of ethical and societal implications of technology development to the fore. Yet, there is little guidance on how to do such research in practice, let alone how to innovate in ethically and socially sound ways. This paper addresses these issues in the context of a large scale EU funded project developing system of system innovations in IT supported emergency response. Building on collaborative design and a range of other approaches, the paper argues that just like ?usability?, ethics cannot be invented or decided by experts, but has to be the product of engagement with the technology by directly or indirectly implicated publics. Facilitating such publics is a central element of what we call ?ELSI Co-Design?. The paper outlines the theoretical and methodological underpinnings of this approach.

Abstract: Warning systems are technically, socially, and organizationally shaped and rest on specific assumptions concerning human behavior during disasters. The common notions about people?s behavior in disaster situations are often not based on empirical data, but rather on so-called ?myths? which overemphasize rare and situation-dependent extreme behaviors such as panic, disaster shock, looting or helplessness. Due to the fact that these expectations are shaped within social environments, different stakeholders such as a heterogeneous population and professionals exhibit different assumptions. These assumptions may not only be misplaced, they additionally interfere with warning systems. The paper compares empirical results of three connected surveys: a comprehensive document analysis on disaster behavior, qualitative interviews with disaster relief workers and a quantitative representative poll. By contrasting the status of research with professional narrations as well as with the people?s expectations, different expectations and their variations are explored.

Abstract: As growing emphasis is placed on engaging with big ?crisis? data, including data from social media, GPS, and satellite, adequate policies and measures must be in place in order to use this data in an ethically and legally responsible manner. The current working paper introduces the BYTE study, which is working towards identifying and understanding the various positive and negative externalities, or impacts, associated with the use of big ?crisis? data. This insight paper provides a preliminary discussion of various externalities that may be encountered in this study. By doing so, the authors highlight the need for additional research in this area to promote ethically and legally responsible crisis data practices.

Abstract: This paper explores the added-value of using interactive head-mounted displays to support command and control of first responders during emergency response. Specifically, it describes and evaluates a prototype system that makes use of Google Glass to enable in-field receiving of information from a command center, as well as in-field navigation and video streaming. The viability and usefulness of the concept was evaluated through a set of end-user workshops and interviews. A small-scale experiment was also conducted to assess the efficiency of using head-mounted displays for in-field navigation, as compared to handheld devices. Findings from workshops and interviews suggest that head-mounted displays could be a valuable supplement to radio communication, with potential for reducing information misinterpretation, and for enhancing information quality. Results from the experiment indicate that head-mounted displays have the same level of efficiency as handheld devices when used for basic navigation tasks.

Abstract: When discussing human casualties from a severe earthquake with regard to urban disaster mitigation planning, it is important to clarify the characteristics of the spatiotemporal distribution of people. In this paper, we construct a model that estimates the spatiotemporal distribution of automobile users using data from the Person Trip Survey and the Road Traffic Census. We use this model to estimate the spatiotemporal distribution of automobile users in Tokyo and demonstrate several ways to apply this data to urban disaster mitigation planning.

Abstract: In the event of a disaster, high resolution knowledge of expected population distribution is a boon to the situational awareness of disaster managers and first responders. Knowing the expected locations of large throngs of people can greatly affect distribution of aid and response infrastructure. Effective dissemination of this information can be realized by using a myriad of readily available technologies.
With the modern proliferation of smart phones, pervasive Internet and freely available social media applications, population distribution can be estimated from the constant aggregation of crowd sourced data. Twitter and Instagram both publish geolocated data, which is then processed by a cloud-based, enterprise application to generate heat maps. The heat maps are then shown in a real-time geographic information system that is visible to any mobile device with a web browser.

Abstract: Given the importance of geographic information for riverine flood evacuations, a geographic information system (GIS) is a vital tool for supporting successful flood evacuation operations. This paper discusses the development of a GIS-based riverine flood evacuation model which used to model trip distributions between flooded areas and relocation shelters. As the ultimate goal of this research is to simulate, model, and optimise a planned evacuation, all components of evacuation time have been considered (e.g., travel time between flooded areas and relocation shelters, warning time for each flooded area, and the time needed for evacuation before these areas get inundated). As well, variation in population (static and dynamic population) within the flooded areas has been considered.

Abstract: In a large earthquake, rescue operations and fire-fighting are obstructed by fire-spreading and street-blockages. Therefore, it is important to quickly collect and utilize disaster information for disaster mitigation. In this paper, firstly, we develop a Web application for posting and viewing information collected by users in real time. Using this system, it is possible not only to easily share disaster information among users but also to apply to damage forecast such as fire-spreading. Next, we demonstrate the usefulness of the Web application by the following evaluation view points: (1) relationship between the access time of emergency vehicles from fire stations to the locations of fires and the ratio of collected information on street-blockage which is assumed to be collected with this system; (2) reciprocating time between a server and a client which is dependent on the number of users and band limitation after the occurrence of a disaster.

Abstract: Over the last couple of years Volunteered Geographic Information (VGI) and particularly OpenStreetMap (OSM) have emerged as an important additional source of information in disaster management. The so-called OSM Crisis Maps are primarily developed by OSM contributors who work remotely. While local OSM contributors know their area of interest and rely upon local knowledge, often the sole basis for the remote mapping is satellite imagery. This fact may raise doubts about the quality of the Crisis Maps. This study introduces an experimental approach to assess the data quality that remote mappers produce. In an experimental setting, data sets produced by a group of remote mappers are evaluated by comparing them to data sets created by a selected expert mapper with local knowledge. The presented approach proved to be useful for assessing data quality of remote mapping and can be used to support decisions about the suitability of crowdsourced geographic data.

Abstract: Maps, showing the tactical or the administrative situation at any particular time, play a central role in disaster management. They can be realized as interactive map mashups. In addition to classical information (weather, water levels, energy network, forces), they can also be used to present a view on citizen-generated content, e.g. from social media. In this paper we offer insights into how mobile map mashups can assist citizens during infrastructure failures that often occur in large-scale emergencies. Based on a review of approaches and mobile applications from literature and especially from practice, we present SOMAP (social offline map), a mobile app we developed in Android. It offers offline map functionality in terms of (A) pro-active loading and storing of potentially needed maps of the respective area as well as (B) the possibility of exchanging information from social media using Bluetooth. The application was evaluated qualitatively, to gain insights into the potential of such applications.

Abstract: Crisis management training software is gaining researchers? as well as practitioners? interest. In order to truly support organizations it is important that such software responds to actual user needs. The aim of this study is to compare existing initiative described in research with the needs of the users and to identify possible research directions for forthcoming studies. The literature review shows that discussions on users? needs are superficial at best. The software described in research mainly focus on co-located execution of exercises, often in the form of simulations. Furthermore, a ?right or wrong? behavior is usually built-in. Empirical data from a web survey indicate that flexibility in time and space during training is a fundamental user need. This is not particularly acknowledged in earlier research. Neither is the users? wish for better support in designing exercises. We propose that system flexibility, modularity and pedagogy for computer based crisis training are urgent issues for future research.

Abstract: Smartphones are becoming increasingly useful in disaster management both to provide useful information to victims and to coordinate relief operations. However, a lack of technological expertise as well as considerable amount of time and cost required to build mobile applications prevents the rapid deployment of useful applications by humanitarian organizations for different crises. In this paper, we describe a participatory design workshop that we conducted at the International Committee of Red Cross to identify challenges of adopting mobile technologies within relief organizations. Through this workshop, we identified major challenges associated with developing mobile applications: lengthy development and deployment cycle, costly budget, and frequent requirement changes. We then introduced our framework that enables non-programmers to quickly develop and deploy mobile applications in these situations. The workshop participants identified three areas where our framework improved upon existing mobile solutions: reducing data integration overhead, fast prototyping for app development, and customization of apps.