Copyright. Citation Information: Noran, O., Effective Disaster Management: An Interoperability Perspective. Lecture Notes in Computer Science, 2011. 7046: p. 112-121.
Effective Disaster Management: An Interoperability Perspective Ovidiu Noran Griffith University Australia, School of ICT
[email protected]
Abstract. Natural and man-made catastrophic events appear be steadily increasing in intensity and frequency. Proper preparation, response and recovery are essential if mankind and its vital systems are to cope with and survive large-scale disasters. Unfortunately, on account of historical and geographical reasons, the organisations responsible for delivering emergency response services often under-perform (or even fail), typically due to a lack of proper interoperation and collaboration. This paper attempts to analyse interoperability issues specific to disaster management and propose a way forward based on the advances in interoperability research, using an enterprise integration and architecture perspective in order to ensure a holistic approach. Keywords: Interoperability, Emergency Response, Enterprise Integration
1 Introduction Mankind appears to rely on increasingly complex systems in order to survive and evolve. At the same time, the rate and force of natural and/or man-made disasters (some triggered / augmented by climate change) appears to be on the rise. In this context, it is now more urgent than ever to be able to promptly and effectively prevent, prepare for, respond to and recover from catastrophic events. Governments typically put in place policies and organisations in order to address disaster management aspects. Of course, such disaster management organisations (DMOs) exist in a multi-dimensional continuum (time / life history, geographical location, level of governance etc), that inherently results in their independent evolution and heterogeneity. While generally organisational diversity is not a detrimental aspect, it does require additional effort towards proper and effective collaboration [1]. Coping with large scale catastrophic events typically requires resources and capabilities beyond those of any individual organisation, making effective cooperation and interoperation essential. As has been seen on many unfortunate occasions, failure to achieve this in the case of emergency management organisations is likely to have dire consequences including preventable loss of property and human life. This paper aims to analyse disaster management–specific interoperability problems and to propose ways to address them using the knowledge accumulated in state-ofthe-art interoperability research. The analysis and proposals will be performed from an enterprise integration and architecture point of view promoting a holistic approach through a combination of frameworks that aims cover all aspects deemed as relevant.
2 Current Issues in Disaster Management The operation of Emergency Services is typically legislated at state, national and international levels (see [2-5]). Unfortunately however, merely instructing organisations to cooperate using high-level generic directives does not bring true collaboration and / or interoperability. The consequences are extended response times, confusion on the situation on the ground, dispute / competition as to who, where and when is in charge, difficulty in coordination with other teams’ systems due to incompatibilities in infrastructure and difficulty in filtering / validating the flood of information generated during disaster events. For example, lack of consistency in the alert notice type and delivery format may delay warnings or flood the population with ambiguous / irrelevant messages [6, 7]. This leads to sub-optimal preventative action and response by the intended recipients and potential property and life loss. Two main approaches are used to address the above-mentioned problems; they involve either centralisation (hierarchical command) or federalisation of emergency services. Irrespective of the approach used however, proper emergency response and cooperation has still not been achieved, as reflected in criticism by authors and community [8-10]. The use of military operations-style network-enabled capabilities as the backbone of disaster management [11] is valid only as part of the overall disaster management effort (e.g. during some life cycle phases of the disaster events). It appears that in fact the root causes are the inadequate information and knowledge flow and quality between the participants [12, 13], the lack of trust, organisational confusion and competition fallacies. True and efficient collaboration is not possible unless the organisational cultures, processes and resources of the participants possess the required interoperability preparedness [14]. Another aspect that appears to have been addressed to a lesser extent is the life cycle of the DMOs, task force, other controlling entities (e.g. laws, standards etc) and the disaster event(s). The interoperability requirements will vary in this context. Analogies with other domains may help provide potential solutions. For example, the DMOs’ situation resembles that of commercial enterprises that need to cope with a global business environment requiring them to tackle projects beyond their own resources and knowledge. A typical response to this problem in the commercial domain is to set up or join Collaborative Networks (CNs) that act as ‘breeding environments’ [15] for Virtual Organisations (VOs – groups of companies that act as one for a limited time) and can be promptly created in order to bid for, win and complete projects requiring combined resources and know-how. Another analogy is that of allied armed forces that need to cooperate in crisis situations. Standardised agreements on language and systems and joint exercises are used in this case [16]. DMOs are situated somewhat in between these two cases as they may have full time / militarised and voluntary / reserve staff components, depending on the geographical location and local legal and administrative situation. Thus, successful concepts from the commercial and military areas can be reused provided due customisation is performed so as to fit the specific DMO setting. The ‘communities of engineering practice’ concept described in [17] can also be applied to emergency management so as to facilitate experience sharing, collaborative learning, professional development and to kick start transnational and trans-cultural partnerships.
3 Interoperability Frameworks Interoperability is understood here as enabling the exchange of information and its use but also as ability to perform a function on behalf of another entity [18, 19]. Previous research and practice have identified various aspects of interoperability. For example, DoDAF (Department of Defence Architecture Framework) [20], via its Levels of Information System Interoperability (LISI), defines manual, peer-to-peer, distributed, integrated and universal levels (in fact expressing a kind of interoperability maturity). The European Interoperability Framework (EIF) [21] defines organisational, semantic and technical interoperability. The Athena Interoperability Framework (AIF) [22, 23] defines data, processes, services and business. IDEAS (Interoperability Development for Enterprise Applications and Software) [24] defines three levels: data, application and business process. Whitman and Panetto [1] identify technical, syntactic, semantic and pragmatic interoperability levels. It appears that various frameworks assign different meanings to the same concept (e.g. level) and different names to the same concept (e.g. aspects / viewpoints). Some concepts appear to be orthogonal, as illustrated in Chen’s [19] framework based on EIF, IDEAS and ATHENA, which is also adding the concept of barriers. Some frameworks combine concepts (and approaches) in matrices or cubes. The following sections will attempt to employ the concepts / frameworks that appear to be most relevant / applicable to disaster management.
4 Disaster Management Interoperability Reasoning about improving DMO interoperability raises several questions. To what extent is interoperability required? What components / aspects of the participants need to interoperate? How can be ensured that all necessary aspects are covered and interoperability is preserved over time as all participants evolve? As each disaster event is unique, there is no ‘optimal’ DMO interoperability level to fit all crisis situations. At a minimum, the participating organisations should display compatibility, so at least they don’t hinder each other’s operations (see Fig.1). Full integration may not be desirable in this case as it could imply that the DMOs cannot function independently at full capacity. In an emergency situation, it is quite possible that one or several participants (or components thereof) could be affected and may even cease to function; the rest must be able to continue without significant performance loss (e.g. similar to the ARPANET resilient network concept [25]). Resilience and agility are essential in this case. Even if a central point of command (Emergency Command Centre) was secure and unaffected by the disaster event(s), the coordination provided by it could be severely affected by communication breakdown. The organisations involved should be able to continue within acceptable operating parameters in such scenarios.
Minimal (Compatibility)
Interoperability level
Full Integration
Fig.1 Desired interoperability level range (based on [1]) The suitability of an interoperability improvement approach must be analysed in view of the framework components presented in the previous section. ISO14258 [26] establishes several ways to achieve interoperability: integrated (common format for all models), unified (common format at meta level) and federated (no common format, with participants negotiating an ontology on the fly so as the meaning of models is understood the same way). In the case of DMOs, it appears that the unified approach is most suitable as both full integration and federalisation did not seem to succeed and achieve the desired results due to organisational heterogeneity of DMOs and the impossibility to properly negotiate in the limited time available in the case of a disaster event. The unified approach assumes that an ontology is negotiated in advance. This requires that the DMOs spend time together prior to disaster events in order to agree on meanings associated with the concepts used to exchange knowledge. Once this is achieved, the semantic requirements for proper interoperability should be promptly met in the task forces formed by the participant DMOs. The technical aspect of interoperability (and associated barrier [19]) appears to be considered the most straightforward to solve. However, disaster management task forces may comprise interstate/international organisations relying on different infrastructure and thus technical details may become the stumbling block if not given sufficient prior attention. This may relate to both software and hardware aspects. The pragmatic aspect of interoperability as defined in [1] relates to the willingness and capacity of the participants to interoperate. Although this aspect is mandated by governments, the human side needs attention prior to task force formation as to allow gaining trust and knowledge of the other DMOs. The syntax interoperability aspect applied to DMOs refers to common machine but also human data exchange formats – for example the adoption of English as the language spoken by an emergency task force comprised of international DMOs. This appears to be reasonably easy to achieve; however, syntactic interoperability does not solve the semantic aspect, i.e. correctly understanding the intended meaning of the information exchanged. In the case of humans, non English-speaking background crews may understand or express concepts in an unintended way – which in the case of EMOs may have tragic consequences. ‘Cultural interoperability’ [1] - whether organisational, local or national - is one of the hardest obstacle to overcome. The only current true solution in the author’s opinion is regular immersion of the participant organisations in each other’s cultures. This will facilitate the transfer / conversion of tacit and explicit knowledge between the participants, which is one of the most important goals of interoperability. Although cultural interoperability is not limited to language, an analogy can be made to human language translators; thus, an accurate translation can only be performed by a person that has a thorough knowledge not only of the languages (and associated
jargons) but also of the philosophy / cultures of the nations speaking those languages. A good translator would have ideally lived for some time in all countries speaking the languages they attempt to translate. Within the universe of discourse of a company, the ‘translators’ could be the enterprise architects who are required to speak the technical jargon but also understand the culture of all departments (e.g. business, management, engineering and information technology). A good enterprise architect would have multiple qualifications and spend time in each department thus gaining a deep understanding of all company areas, being able to bridge them and thus assist the enterprise integration effort. Organisational interoperability is an important aspect in disaster management as participants may often have different organisational and management structures. The issues identified by Chen [19] based on the EIF framework, namely responsibility, authority and type of organisation can all impact heavily on the functionality of the disaster management task force. In a crisis situation, the roles (mapping of the human resources onto the decisional structure) and hierarchy must be clear to everyone from the start so that the task force can focus on managing the disaster event rather than spend critical time figuring out its own modus operandi (who does what, who is in charge etc).
Approach Barrier Business Service Process
Integrated Unified
Data
Federated
Concern Conceptual
Organisational Technological
(syntax + semantics)
Legend: = essential
= tertiary
= secondary
= n/a
Fig. 2 Interoperability issues and approach in disaster management (based on [19]) Using the interoperability concerns identified by the Athena project, the data and process aspects appear to be the most relevant in disaster management. Thus, the ability of task force participants to extract and exchange data from potentially heterogeneous data repositories is paramount to being aware of the conditions on the ground and to avoid sending personnel into an unknown and potentially lifethreatening situation. The high volume and erratic reliability of data present in a disaster event will compound this problem. Prior agreements on data format and especially on meaning are essential. Process interoperability in the author’s opinion
concerns both the capability to perform joint operations - which underpins the entire task force concept – but also to ‘take over’ and perform a process instead of a participant that may have been temporarily or permanently disabled. Again, prior agreements but also joint practice on the ground should be performed in order to address this important interoperability concern. The service aspect of interoperability concerns the possible interconnection of the services provided by the DMOs so as to form a more complex service and involves syntactic and semantic aspects that should be solved well in advance to the formation of a task force. The business interoperability aspect refers to organisational culture and management style of the interacting DMOs (addressed in the cultural interoperability section above). Fig. 2 shows the proposed relevance of interoperability aspects in the case of disaster management using the framework presented in [19]. The framework has been selected because it combines concepts of several other interoperability frameworks. Note however that it is not ‘complete’ - while cultural interoperability may be partly represented in the organisational aspect, other important aspects, such as life cycle are not adequately represented. Typically, in order to ensure that all necessary interoperability aspects are covered, it may be necessary to use a combination of frameworks. Previous research has shown that the modelling framework of a generalised (i.e. abstracted from several others) architecture framework can provide a more complete viewpoint repository that can be used to model and select relevant interoperability aspects for a specific project type, including disaster management.
4 Enterprise Architecture, Life Cycle and Disaster Management The tasks involved in reaching and maintaining EI are on-going, complex, and often involve ill-defined constraints and uncertainty. Enterprise Architecture (EA) holds the promise to help stakeholders manage the EI effort in a consistent and coherent fashion, so as to achieve and maintain integration within and between organisations. For this reason it has been considered that the use of EA artefacts and methods can complement the improvement of EMO interoperability. To illustrate this point, we have selected ISO15704 Annex A (GERAM1, see [27] for details), a generalised enterprise architecture framework (AF) based on several mainstream AFs. Among other advantages, the modelling framework (MF) of GERAM’s reference architecture (GERA) allows representing all other aspects in the context of life cycle. The left hand side of Fig. 3 shows how the typical disaster management activities [3] can be mapped to the life cycle phases of a disaster event (represented using the GERA life cycle dimension). Further on, in the same figure it can be seen that the relevance and applicability of the aspects proposed by the reviewed interoperability frameworks are in fact depending on the specific life cycle phases of the disaster event and the DMOs’ actions to address them. Therefore, the use of AF artefacts does promote and enrich the understanding of the interoperability problems specific to disaster management. Further mappings have been performed, involving other aspects
1
The Generalised Enterprise Reference Architecture and Methodology
of the GERA MF. They are not shown here due to space limitation constraints and will be disseminated in other papers. Disaster Interoperability Disaster Event Management Aspects / Importance Life Cycle Actions Mgmt.
Conceptual
(syntax | semantics)
Identification
Technological Organisational
Prevention
Concept
Business
Requirements Prel. Design
Preparation
Det. Design Implementation
Process
Response Data
Operation Decommission
Service
Recovery
Unified Approach
Fig. 3 Disaster event mapping and modelling using a GERA-based formalism Reviewing the current interoperability and disaster management body of knowledge it has become clear that while at some levels interoperability can be promptly and significantly improved, the conceptual (especially semantic side) and human-related interoperability aspects such as trust, management, organisational and local culture must be addressed in advance. In addition, organisations evolve and thus the solutions found must be regularly reviewed. For this reason, the life cycle and management vs. mission accomplishment views of GERA have been employed in Fig. 4 in order to a) illustrate and propose a solution in the form of a ‘breeding environment’ approach and b) analyse the interoperability requirements in this context. Thus, Fig. 4 shows how DMOs can form a so-called ‘disaster management breeding environment’ (DMBE). This structure allows participants to work together (whether in same space and/or time or not) in order to iron out critical interoperability issues well in advance, rather than wasting precious time during disaster events. For example, agreements can be reached and interfaces built if required to bridge technological and technical aspects such as infrastructure gaps. The conceptual syntactic interoperability can be addressed by agreeing on and upholding standardised formats. Semantic interoperability can effectively be addressed in this case by the personnel of participating DMOs spending time together physically (in the same location / time) and immersing in each other’s cultures as previously described in this paper. This immersion is seen by the author as the only currently feasible way to also address the organisational, business and cultural interoperability, which together with the semantic aspect represent the ultimate goal of the entire interoperability effort. The arrows between the entities shown in Fig. 4 represent amalgamated interoperability requirements, (detailed in further diagrams and omitted due to space limitations) in respect to importance, aspects, levels etc identified in interoperability frameworks. Thus, for example ‘interoperation’ of the population POP with the Government Govt is important as may result in changes to legislation and DMO and
DMBE interoperability requirements. However, interoperability between DMOs within the task force DMTF and between the DMTF and POP is paramount as it will directly decide the amount of lost property and casualties. Past (and sometimes tragic) experience has shown that the two major goals of the disaster management interoperability improvement effort partly reflected in Fig. 4 should be a) whether POP receives, understands, believes and uses DMTF warnings and directives and b) that DMTF participants can properly interoperate during the disaster event DE. P
M
Id C R
POP
PD
EML
Legend:
DD I Op D
Govt DMO
DMBE
M
EMC
DMO: Dissater MgMt Orgs; DE/JE: Disaster Event / Joint Exercise DMBE: Disaster Mgmt Breed Env EML: Emrg. Mgmt Laws DMSt: Disaster Mgmt Standards EMC: Emrg. Mgmt Contractors; TFRM: Task Force Reference Model DMTF: Disaster Mgmt Task Force Gvt: Government POP: population : Operation phase, Production & Management : Operation Phase, Production : Possible scenario
TFRM DMTF
DMSt
DE / JE
Life cycle phases: Id: Identification; C=concept; R=requirements, PD=preliminary design, DD=detailed design, I=implementation, Op=operation, D=decommissioning. Other aspects: P=Production / Service, M=management
Fig. 4. Interoperability in the context of a Breeding Environment and Task Force solution
The validity and effectiveness of the DMBE concept can be tested by joint exercises (JEs) simulating DEs; this will also allow practical and perhaps still uncovered interoperability problems to surface and be tackled. Importantly, the resilience of DMTFs created can be also assessed in JEs through various scenarios. Corrective action can then be taken to improve preparedness towards real crisis situations.
5 Conclusions and Further Work Disaster management interoperability related issues are best addressed in advance and in a holistic manner so that when a disaster event occurs an efficient, collaborative task force can be promptly put together. Using state-of-the-art research results in
interoperability and enterprise architecture, the paper has analysed interoperability issues specific to disaster management and has reached several conclusions. Thus, the interoperability level of DMOs should best not reach full integration so as not to affect resilience. The cultural / organisational and semantic aspects of interoperability, as the most important and difficult to solve, should be tackled in advance, allowed time and be periodically reviewed. The life cycle of the participants, task force and disaster event(s) must form the background context to the entire interoperability improvement effort. It is also proposed that the adoption of breeding environments and joint exercises concepts customised for disaster management will greatly assist with all the interoperability aspects described in this paper. Further research will concentrate on testing and validating the findings with CMO stakeholders within several case studies.
References
1.
2.
3.
4.
5.
6.
7.
8. 9.
Whitman, L. and H. Panetto, The Missing Link: Culture and Language Barriers to Interoperability. Annual Reviews in Control, 2006. 30(2): p. 233-241. Government of South Australia. Emergency Management Act 2004. 2004 [cited 2011 Mar 30]; Available from: http://www.legislation.sa.gov.au/LZ/C/A/EMERGENCY MANAGEMENT ACT 2004.aspx. Australian Government. Attorney's General's Office - Emergency Management in Australia. 2011 [cited 2011 Mar 30]; Available from: http://www.ema.gov.au/. Federal Emergency Management Agency. National Response Framework. 2011 [cited 2011 Mar 30]; Available from: http://www.fema.gov/pdf/emergency/nrf/about_nrf.pdf. United Nations International Strategy for Disaster Reduction Secretariat (UNISDR). Hyogo Framework for Action 2005-2015: Building the resilience of nations and communities to disasters. 2011 [cited 2011 Mar 30]; Avail: http://www.preventionweb.net/files/1037_hyogoframeworkforactionenglish Queensland Floods Commission of Enquiry. Submissions to the Enquiry. 2011 [cited 2011 July]; Available from: http://www.floodcommission.qld.gov.au/submissions/. Victorian Bushfires Royal Commission. Submissions to the Enquiry. 2009 [cited 2011 July]; Available from: http://www.royalcommission.vic.gov.au/Submissions/View-Submissions. Clark, J.L., Practical aspects of federalizing disaster response. Critical Care, 2006. 10: p. 107-113. Wiese, C.R., Organizing Homeland Security after Katrina: is adaptive management what’s missing? Public Adm Review, 2006. 66(3): p. 302-318.
10.
11.
12.
13.
14.
15.
16.
17.
18. 19.
20. 21.
22.
23.
24. 25. 26.
't Hart, P., et al., The Politics of Crisis Management: Public Leadership under Pressure. Journal of Homeland Security asnd Emergency Management. 2005, Cambridge UK: Cambridge University Press. von Lubitz, D.K.J.E., J.E. Beakley, and F. Patricelli, Disaster Management: The Structure, Function, and Significance of Network-Centric Operations. Journal Homeland Security and Emergency Management, 2008. 5(1): Art 42. Wickramasinghe, N. and D.K.J.E. von Lubitz, Knowledge Based Enterprises: Theories and Fundamentals. Journal of Homeland Security and Emergency Management. Vol. 5. 2007, Hershey PA: IGP Publishing Group,. Prizzia, R. and G. Helfand, Emergency preparedness and disaster management in Hawaii. Disaster Prevention and Management, 2001. 10(3): p. 163-172. Kapucu, N., T. Arslan, and F. Demiroz, Collaborative emergency management and national emergency management network. Disaster Prevention and Management, 2010. 19(4): p. 452-468. Camarinha-Matos, L., et al., Collaborative networked organizations Concepts and practice in manufacturing enterprises. Computers and Industrial Engineering, 2009. 57(1): p. 46-60. NATO. Interoperability for joint operations. 2006 [cited 2011 July]; Available from: http://www.nato.int/docu/interoperability/interoperability.pdf. Molina, A., J. Aca, and P. Wright. Enterprise integration and networking: issues, trends and vision. in Knowledge sharing in the integrated enterprise interoperability—strategies for the enterprise architect (proceedings of ICEIMT04). 2004. Toronto, Canada: IFIP. Vernadat, F., Enterprise Modelling and Integration: Principles and Applications. 1996, London: Chapman & Hall. Chen, D. Framework for Entrerprise Interoperability. 2006 [cited 2011 July]; Available from: http://www.fines-cluster.eu/fines/jm/Downloaddocument/53-Framework-for-Enterprise-Interoperability-Chen.html. DoD Architecture Framework Working Group. DoD Architecture Framework Ver 1.0. 2004 [cited 2007 Feb 2007]. EIF, European interoperability framework for pan-European eGovernment services. Luxembourg: Interoperable Delivery of European eGovernment Services to public Administrations, Businesses and Citizens (IDABC). 2004. ATHENA State of the art of Enterprise Modelling Techniques and Technologies to Support Enterprise Interoperability. Deliv D.A1.1.1 2004 [cited 2011 Mar 30]; Available from: http://www.athena-ip.org. Berre, A.-J., et al. The ATHENA Interoperability Framework. 2006 [cited 2011 Jul]; Available from: http://www.txtgroup.com/newsletter/attachment/Athena_Paper.pdf. IDEAS. IDEAS Project Deliverables (WP1-WP7), Public reports. 2003 [cited 2011 Jul]; Available from: www.ideas-roadmap.net. Heart, F., et al., ARPANET Completion Report, Bolt, Beranek, and Newman, Editors. 1978: Burlington, MA. ISO, ISO14258 Industrial Automation Systems – Concepts and Rules for Enterprise Models. 2005.
27.
ISO/IEC, Annex C: GERAM, in ISO/IS 15704:2000/Amd1:2005: Industrial automation systems - Requirements for enterprise-reference architectures and methodologies. 2005.
