Risk perception–based post-seismic relief supply allocation in the Longmen Shan fault area: Case study of the 2013 Lushan earthquake

Use of OR in earthquake operations management: A review of the literature and roadmap for future research

“…Interestingly, simulation, which is used frequently in pre-earthquake stages (mitigation and preparedness), has rarely been used in post-earthquake response problems. Heuristic [167,168, Simulation [171,205] Machine learning [153] Decision analysis [171] Soft OR [206] Road damage assessment Machine learning [207] * References highlighted in bold incorporate more than one methodology and/or address multiple problem types.…”

“…Four other methods besides mathematical programing and heuristics have been applied to relief distribution. This includes: (i) a system dynamic model to analyze a relief distribution system built for the Longmen Shan fault, China, where many destructive earthquakes have occurred [205]; (ii) Soft OR for developing a conceptual model of post-disaster survivor perception-attitude-resilience relationships to inform emergency logistics operations in a way that takes into account perspectives of both government planners and the psychology of affected populations; (iii) machine learning (neural networks) for designing an efficient blood supply chain [153] and predict the structural status of road links when deploying relief [207]; and (iv) decision analysis to assess performance of relief distribution based on demand coverage, logistics costs, and response time [171].…”

“…Contrary to this, we observed that only 138 of 211 studies in our review present a problem definition expressly focused on earthquakes and EOM decision making (see Appendix B). In only a few studies were emergency operations and stakeholder roles defined by field experts [122,205]. Interestingly, studies addressing reliability and vulnerability analysis and integrated disaster management have gone the furthest in terms of using earthquake-oriented problem specifications.…”

“…There is broad recognition that in order to achieve buy-in and a measurable improvement in EOM performance, all key stakeholders need to be involved both in implementation and problem identification and modeling (not necessarily fine-grain details but at least general structure) [144,157,176,205]. Lack of stakeholder involvement in model conceptualization and development often leads to more theoretical and less realistic problem definitions, case studies that provide limited insights, and ultimately low likelihood of proposed methods ever being implemented [17].…”

“…key performance indicators) of their decision support models [61,64,65,122]. Finally, a third group mainly looked at post-disaster psychological metrics of local residents directly impacted by large-scale earthquakes based on interviews and questionnaires [205,206]. Only one study addressed a problem proposed by EOM practitioners themselves [107].…”