A Multi-stage Stochastic Programming for Redesigning the Relief Logistics Network: A Real Case Study Hamidreza Rezaei* Department of Industrial Engineering, Yazd University, Yazd, Iran +982166803083 Hamidreza.rezaei21@g mail.com Hasan Khademi Zare Department of Industrial Engineering, Yazd University, Yazd, Iran +983531232423 hkhademiz@yazd.ac.ir Mahdi Bashiri Department of Industrial Engineering, Shahed University, Tehran, Iran +98 3319118651 bashiri.m@gmail.com ABSTRACT Disasters inevitably cause sweeping problems affecting all surroundings. Therefore, it seems that right decision making for agile and efficient management in disaster management is not negligible. Due to designed network in the past, current entities, facilities and network may not be optimal and to a great extent usable. In this paper a comprehensive model and solving approach have been proposed to redesign the relief network in dealing with the preparedness and response phases. In this regard, previous investigations on the preparedness phase have often been limited to the location of eligible facilities without considering other important factors such as current and working assets, entities and configuration. Thus, the present study proposes a reconfiguring and repositioning model in order to simultaneously assess whether existing distribution centers should remain, be consolidated or phased out as well as whether new facilities should be established and subsequently supply and demand requirements consideration. Moreover, in the proposed model, multi-stage stochastic programming has been implemented on a real data gathered by ArcGIS Software and demand scenarios derived from relevant references. The results based on a real case study in Tehran indicate definite advantages in the re-positioning or reconfiguring model compared with current configurations. CCS Concepts • Information systems~Enterprise applications Keywords Disaster Management; Relocation; Redesign; Preparedness Facility; Multi-stage Stochastic Programming; GIS Data. 1. INTRODUCTION Disasters have widespread effects on a diverse range of our surrounding. In terms of terminology, disasters are classified in two group including natural (e.g. earthquakes, floods, hurricanes, tsunamis) and man-made (war, political/tribal disturbance, famine). Moreover, the measures for soothing the disasters’ outcomes are also considered in some phases: pre-disaster such as mitigation, preparedness and after disaster such as response and recovery phases. The main objects in disaster management are the reduction of injuries and fatalities rate for humanity purposes and protection of vital infrastructures (i.e. links, bridges, facilities and etc.) against destruction and damage. The decisions in mitigation and preparedness are taken as preventive measures to support the further stages and phases such as response and recovery. Peeta et al. [1] have investigated the best choice for investment in a long- term on strengthening the network’s links with the intention of more accessibility and connectivity. On the other hand, the decisions taken in preparedness phase (locations, relations between echelons, capacities and etc.) facilitate the relief in the response phase (for more realization see the investigation of Beraldi and Bruni [2] as a case in point). In the previous papers in the preparedness phase, the main attentions have been focused on the design a new network based on predicted parameters but in this investigation, it has been supposed the relief network that has been designed should be redesigned in order that the post disaster cause less fatalities and damages. In this regard, It seems that the proposed reconfiguration model should be able to consider some vital issues about the location of distribution centers (DCs) including “which facilities should remain, established, phased out or consolidated?”. Moreover, it should determine the relations between echelons as well as quantity of relief goods shipped from the suppliers to DCs and then to demand nodes. Finally the proposed model and its solving approach should be able to compare the performance of the reconfigured network and the current configuration during the disaster occurrence in terms of fatalities, costs, shortages, covered demands and any factor that can be vital in decision-making. The rest of the paper is organized as follows. The literature of the relief network design in the preparedness phase is discussed in section 2. In the next section the problem description is argued, and then the proposed formulation of re-designing model with consideration of scenarios for a path-in-the-scenario-tree-based formulation are presented in section 4. In section 5, a real case study for a specific district of Tehran is implemented and it will indicate that to what extent the proposed approach works applicable. Thereafter, the paper ends with some conclusions and relevant future research ideas in section 6. 2. LITERATURE REVIEW The distribution of published papers in preparedness phase reveals that this phase has taken considerable attention in the recent Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from Permissions@acm.org. ICIME 2017, October 9–11, 2017, Barcelona, Spain © 2017 Association for Computing Machinery. ACM ISBN 978-1-4503-5337-3/17/10…$15.00 DOI: https://doi.org/10.1145/3149572.3149614 M. B. Fakhrzad Department of Industrial Engineering, Yazd University, Yazd, Iran +989131548146 mfakhrzad@yazd.ac.ir 111