Content uploaded by Adam Widera
Author content
All content in this area was uploaded by Adam Widera on Oct 04, 2016
Content may be subject to copyright.
The International Emergency Management Society
2015 Annual Conference, 30th September - 2nd October 2015, Rome, Italy
GDACSmobile - An IT Tool Supporting Assessments for Humanitarian Logistics
Daniel Link1, Adam Widera1, Bernd Hellingrath2, Tom de Groeve3, Gintare Eidimtaite4, Minu
Kumar Limbu5
1 Research Assistants, European Research Center for Information Systems, University of Münster, Germany
2 Full Professor, European Research Center for Information Systems, University of Münster, Germany
3 Senior Scientist, Joint Research Centre, European Commission, Brussels, Italy
4 Associate Humanitarian Affairs Officer, United Nations Office for the Coordination of Humanitarian Affairs,
Geneva, Switzerland
5 Information Management Specialist, United Nations Children's Fund, Kisumu, Kenya
Email: daniel.link@wi.uni-muenster.de, adam.widera@ercis.uni-muenster.de,
bernd.hellingrath@ercis.uni-muenster.de, tom.de-groeve@jrc.ec.europa.eu, eidimtaite@un.org,
minulimbu@gmail.com
ABSTRACT :
A rising number of people is affected by disasters, such as the 2010 Haiti earthquake, the 2013 Philippines
super-typhoon, and the 2015 Nepal earthquake. In the immediate aftermath of a disaster, humanitarian decision
makers have to assure that action is prompt and targeted although confronted with lack of needed information about
the highly complex and dynamic operational context in the affected area. Mobile technology and crowdsourcing have
emerged as technologies that can help supply much needed information. GDACSmobile is a mobile-enabled IT
solution for the assessment of needs, issues of access, infrastructure damage and other cross-cutting operational issues.
In contrast to similar tools, e.g. Ushahidi or KoBoToolbox, GDACSmobile is an integrated solution addressing both
the general public and professional responders, which closes the information cycle between disaster managers and the
affected population. After all, most first responders are local citizens. In this paper, we introduce the context of the
application and its concept, including descriptions of user groups and information flows that enable effective quality
control of information. Subsequently an application scenario based on the 2015 Nepal Earthquake illustrates the value
of using GDACSmobile to involve logisticians and assessment experts to inform and control the processes of data
collection and information analysis, wherein professional responders and citizens act as primary data sources.
KEYWORDS:
Humanitarian Logistics, Assessment, Data Collection, Information Analysis, Volunteered Geographic
Information, Mobile Devices, Crowdsourcing, Social media
1. INTRODUCTION
1.1. Humanitarian Logistics
1.1.1 The role of humanitarian logistics
Humanitarian logistics plays a central role in disaster relief operations of humanitarian organizations, where it
helps to meet beneficiaries’ needs by procuring, storing and transporting goods that are either distributed
directly to the affected population or that support the provision of services, like medical care. While the term
humanitarian logistics seems to emphasize the physical handling of goods, it is an umbrella term referring to
both logistics and supply chain management, which includes tasks for planning operations.
1.1.2 Tasks, decisions and information needs
During the 2015 Nepal earthquake response, Kathmandu international airport became congested and did only
accept airplanes below a certain weight, which led logisticians who had this information to use road
transportation from Delhi in India as an alternative. This illustrates the importance of information about needs,
supplies, resources and other attributes like costs or delivery times to inform logistics planning, i.e. to identify
The International Emergency Management Society
2015 Annual Conference, 30th September - 2nd October 2015, Rome, Italy
and evaluate effective and efficient ways of acting.
1.2. Assessment
Since humanitarian organizations often operate in countries with poor infrastructure and low levels of data
preparedness, and disasters often drastically change the environment, so-called assessments take an important
role in gathering sufficiently timely and accurate information. The aim of assessments is to generate information
serving as a decision base for other functional areas.
1.2.1 Secondary and primary data collection
Assessments usually begin with collecting and analysing secondary data, i.e. data that has been available from
before the disaster, in order to estimate the needs of the affected population as well as operational constraints.
Secondary data exists in many forms, such as generic country overviews (see e.g. the CIA World Fact Book)
that give an overview of demographics and other aspects, or specialized datasets (e.g. the Logistics Cluster’s
Logistics Capacity Assessments) that describe a country’s logistics infrastructure and resources. If available
secondary data is not sufficient to plan operations, primary data needs to be collected from the affected area via
remote sensing, exploratory teams or other methods. In recent years, mobile technology and crowdsourcing
have emerged as new ways of collecting needed data, not only impacting the capabilities of professional
responders but also the role of the affected population.
1.2.2. Professional responders as data source
Professional responders are arguably the most important source of ground truth from an affected region due to
their education, training and experience. Looking at logistics-related information specifically, for similar reasons
field logisticians can be named as the most valuable source for logistics-related information. However, the
limited capacity of individual responders in a highly demanding context requires a concerted effort of
assessment experts, logisticians and other humanitarian workers who may contribute information, in order to
fulfil logisticians’ information needs to the greatest possible extent. While experienced assessment experts are
highly proficient at adapting to and making sense of the overall situation at hand, most responders excel in other,
specific areas of expertise and benefit from purposefully designed forms guiding their observations, like the
Logistics Cluster’s Logistics Rapid Assessment Tool that draws attention to the most critical aspects of e.g.
airports. Although very useful, paper forms have several limitations in terms of processing speed, error rate and
the ability to capture rich data like geo-located images. This is one reason why humanitarian organizations
increasingly look at mobile devices like smartphones and tablets to support data collection, despite such devices
having their own drawbacks, such as the dependence on electric power or access to telecommunication networks.
Eventually, an effective mobile tool needs to be able to accommodate various kinds of professional responders
while operating with an adequate focus. That is, limiting its application to an area where its usefulness can be
expected without over-exceeding its capabilities by ignoring requirements and limitations, e.g. confining
assessment experts to one particular form design.
1.2.3. Local citizens as data source
The affected population does not only take a primary role in humanitarian response as beneficiaries, but local
citizens often act as first responders (e.g. saving most of the people trapped under rubble), possess relevant
knowledge of the area (e.g. about local suppliers) and can provide information about their needs and the
operational environment (e.g. about the state of the road network or the security situation). The increasing
adoption of mobile phones for access to phone calls, SMS text messaging and Internet-based services among
citizens enables humanitarian organizations to communicate with citizens, gaining information that can
contribute to assessments. Since citizens aren’t usually trained as disaster responders and aren’t familiar with
assessment methodologies, their observations should be guided by specific questions without ignoring the
general chatter on public media like online social networks. Furthermore, collected data needs to undergo
review before being released to professional responders to inform their decision-making, in order to prevent
low-quality observations, however well intended, increasing information overload and adversely affecting
The International Emergency Management Society
2015 Annual Conference, 30th September - 2nd October 2015, Rome, Italy
decision-making. These negative side-effects seem to concern practitioners, who haven’t yet seen a convincing
demonstration of the impact of e.g. results from social media analysis, despite visionary technologists and
specialized researchers making strong efforts to advocate the usefulness of information from citizens.
Integrating these various perspectives demands serious consideration of citizens’ input, be it via the harvesting
of social media data or involving people through smartphone apps, while protecting professional responders
from information overload and useless distraction through a suitable review process.
1.3. Background of the project
Our research group from the University of Münster and the Joint Research Centre of the European Commission
aimed at developing a solution that disaster relief professionals and the affected population can use to acquire
and disseminate primary data within the first four weeks after a major sudden-onset disaster. A first design
cycle in 2012 resulted in the solution “GDACSmobile”, which enables disaster management professionals and
the affected population to share their observations from the affected area, both via the GDACSmobile
smartphone application and via Twitter as an exemplary online social network. The app also enables
professionals and citizens to view published observations to gain valuable information for their decision-making.
To ensure the quality of the information that is disseminated through the client application, the solution employs
a concept to moderate incoming reports. A second design cycle in 2014 focused on improving and extending the
existing solution in terms of usability, interoperability and perhaps most importantly the communication
between moderators and users, establishing bidirectional feedback loops.
1.4. Structure of the Paper
In the remainder of this paper, we will describe the GDACSmobile concept, including the definition of user
groups and a description of necessary information flows. Next, we describe an application scenario. Eventually,
we draw conclusions and give an outlook on the tool’s future development.
2. Concept of the Assessment Tool
2.1. User Groups
GDACSmobile distinguishes the following basic user groups.
Moderated users can be assumed to submit observations that require quality control before reaching a
satisfactory level of confidence regarding the information’s relevance, accuracy and timeliness. General
members of the affected population fall into this group due to their limited understanding of assessment
methodologies and information quality. Technically, moderated users submit observations by including
certain hashtags in their Twitter messages, by using the GDACSmobile smartphone app without a
registered user account, or by registering for a user account without being assigned to an organization.
Non-moderated users can be assumed to submit observations of a sufficient quality to inform
decision-making in humanitarian organizations without prior review. They are affiliated with an
organization, having received at least some training and probably having professional experience.
Technically, a non-moderated user works with a registered user account that is associated with at least
one organization.
Moderators are able to control the quality of incoming observations. As such, they need to be skilled
and experienced enough not only to make high quality observations but to judge their quality, how they
fit into the emerging bigger picture, and which questions need to be asked to build an operational
picture. Technically, moderators are non-moderated users with additional privileges for at least one
organization.
Administrators configure and maintain the system by specifying organizations, user privileges and
The International Emergency Management Society
2015 Annual Conference, 30th September - 2nd October 2015, Rome, Italy
disaster-related attributes. Technically, administrators either have special, dedicated accounts that don’t
need to be associated with any organization and can serve as administrators system-wide, or they are
moderators for at least one organization and can serve as administrators for that organization.
2.2. Information flows
Prior to information flowing in GDACSmobile, the system has to be configured, i.e. an administrator needs to
build forms for the relevant disasters (alerts) that are pulled automatically from the GDACS website (optionally
using templates based on previous disasters), optionally specify first organizations that are going to use the
system and create private missions for these organizations, which are private spaces whose content is not
automatically shared with users outside of the mission space. Once the initial configuration is complete,
information may start flowing, as illustrated in Figure 1.
Figure 1. GDACSmobile Information Flows
The process begins with users submitting observations. If a non-moderated user submits an observation via the
app, it bypasses moderation and will be published according to the initial configuration (e.g. only within a
private mission space). If a moderated user submits an observation via Twitter or the app, it will enter a quality
control loop. Therein, a moderator reviews the observation, possibly draws conclusions affecting configuration
(e.g. by extending a particular form with an additional question) and, if needed, inquire further information from
the observation’s author (e.g. to clarify ambiguous phrasing or ask for more details). Once a moderated
observation shows a satisfying quality, it can be published according to the current configuration (e.g. to all
users in the public mission of a disaster or to a limited user group in a private mission space).
3. Application Scenario (Based on the 2015 Nepal Earthquake)
A 7.8 magnitude earthquake strikes Nepal, triggering several avalanches and aftershocks in different parts of the
country. The situation worsens when another earthquake of 7.3 magnitude strikes the country only a few weeks
The International Emergency Management Society
2015 Annual Conference, 30th September - 2nd October 2015, Rome, Italy
later. The United Nations Office for the Coordination of Humanitarian Affairs (OCHA) estimates 8,700 fatalities,
thousands of injured people, and almost 800,000 buildings affected. Neighbouring communities in India, China
and Bangladesh are affected as well, although with lower intensity. When the government requests international
aid and assistance, aid organizations from different countries initiate the delivery of various goods (e.g.
medicine, equipment, vehicles), financial support, and staff to Nepal; intending to provide relief services like
search-and-rescue and medical assistance.
OCHA’s Disaster Assessment and Coordination (UNDAC) teams run a GDACSmobile server. Once the
magnitude of the disaster has become clear and Nepal has requested international assistance, an UNDAC
administrator creates a private mission space for search-and-rescue (SAR) teams and use an existing earthquake
template to rapidly configure data input forms, asking key questions for the rapid assessment phase. UNDAC
utilizes the International Search-and-Rescue Advisory Group (INSARAG) and the Virtual On-Site Operations
Coordination Center (VOSOCC) to notify search-and-rescue teams. Several team members are already familiar
with the GDACSmobile app, as they have used it in previous exercises, and they are affiliated with their
organization in the system. Once they open and log into the app, it downloads the private mission space and data
collection forms, which the team members review to know where they are asked to direct their attention. After
landing at their destination, they submit first observations about the airport and road infrastructure. Since the
team members are trusted observers, i.e. they belong to the non-moderated user group, their observations
become immediately visible to all members of the private mission space. The private mission space is accessed
by an information manager from the Global Logistics Cluster Support Cell (GLCSC), who studies the submitted
observations to update a previous Logistics Capacity Assessment (LCA) and draft a first situation report, which
many logisticians refer to for their planning. Logisticians thus learn that there are too many aircraft targeting
Kathmandu airport and that it is becoming a bottleneck, also because its (previously sub-optimal) condition
worsened after the earthquake, to a degree where it seems like the weight of incoming aircraft will have to be
restricted. This directs their attention to other airfields in the country that can be expected to still have unused
capacity and lets them use airplanes of a smaller size for transportation. As an additional way of circumventing
the emerging issue of airport access, they consider alternative points of entry, like road transportation from
Delhi to Kathmandu. There, responders on some of the first transports use the GDACSmobile app to take
pictures and enter short descriptions of constraints, such as blocked roads and issues with customs at the border.
Soon after the earthquake, OCHA activates the Digital Humanitarian Network (DHN) as the primary interface
towards digital volunteer and technical communities (VTCs), including the Stand-by Task Force (SBTF) that
focuses on the processing of social media messages and the organization Kathmandu Living Lab (KLL) that
engages in crisis mapping. KLL soon emerges as an important actor possessing much needed local knowledge
and the ability to guide volunteer efforts. VTCs don’t only take UNDAC’s information needs, as specified in
GDACSmobile, as valuable input for their own social media collection and processing activities. They also
support UNDAC in spreading knowledge of GDACSmobile, including Twitter access, among local citizens, and
in moderating incoming reports from local citizens. SBTF members thus review incoming observations within
GDACSmobile’s public mission for the disaster, asking authors clarifying questions where necessary and
marking promising observations as accepted, which makes them visible to professional responders. This sheds
further light on the condition of local road infrastructure, which is utilized by KLL to update OpenStreetMap
and by GLCSC to revise their maps of the area.
4. Conclusion
Logisticians need a variety of information, from up-to-date status reports about major airports to road conditions
and customs procedures. Until date, practice focuses on secondary data (like previous capacity assessments) and
field logisticians as sources of information. The advance of mobile technology and crowdsourcing is promising,
however, so that logisticians may benefit from the potential of other humanitarian professionals and citizens
acting as sources of information. Considering these emerging opportunities requires careful integration into
existing work practices though, so that existing workflows and already strained human resources will not be
disrupted by pushing low quality information, increasing information overload and frustrating responders with
what can all too easily be disregarded as a hype with no impact and adverse effects. Utilizing the potential of
mobile technology and crowdsourcing to supply information useful for logistics decision-making requires not
only a clear understanding of logisticians’ information needs and feasible structures and processes to integrate
The International Emergency Management Society
2015 Annual Conference, 30th September - 2nd October 2015, Rome, Italy
new sources of information with existing work practices but also adequate technology that facilitates data
collection and analysis.
We have developed GDACSmobile as a solution that incorporates professional responders and citizens as
primary data sources. The solution utilizes, among other things, a novel user concept and information flows that
incorporate flexible configuration and effective quality control. While being a promising innovation, the system
does still suffer from various limitations that need to be overcome before its practical usefulness can be
effectively demonstrated, which is a necessary prerequisite to adoption in practice.
GDACSmobile still focuses on primary data, although secondary data, such as existing documents or
websites, can be named as the most important source of information in the very first phase of a sudden
on-set disaster.
It is not realistic to assume that assessment experts would review each and every incoming observation
from moderated users. One way to address this issue is to leverage the capabilities of experts, volunteer
communities and automation in unison, e.g. via a moderation hierarchy featuring a machine learning
system filtering information, which is further distilled by volunteers, until experts review the most
promising pieces of information.
Looking at the future of information systems supporting not only data collection and information
analysis but also decision support, particularly planning, it is necessary to adapt the underlying data
model and open interfaces so that 3rd party decision support systems can automatically collect and
integrate generated information into planning and other areas of decision-making.
When these and further adjustments are made, we have reason to believe that the current positive feedback and
interest from researchers and practitioners will further advance towards adoption and positive impact on
humanitarian action.
Acknowledgements
We want to thank the following students at the Research Group on Information Systems and Supply Chain
Management for their contribution: Anton Becker, Carsten Bubbich, Friedrich Chasin, Jonathan Dölle, Jonas
Juchim, Sven Kronimus, Ferdinand Knoll, Magdalena Lang, Stefan Laube, Marius Pilgrim, Philipp Saalmann,
Mohamed Junaid Shaikh, Yannic Schencking, Martin Vanauer, and Patrick Vogel. For sharing their views in
many discussions during design and development, we also want to thank practitioners Minu Limbu (UNICEF
Kenya), Gintare Eidimtaite and Thomas Peter (both UNOCHA) as well as researchers Tom de Groeve and
Alessandro Annunziato (both JRC). The research leading to these results has received funding from the
European Community's Seventh Framework Programme (FP7/2007-2013) under Grant Agreement n°607798.
