Search results

The recent COVID-19 outbreak and severe natural disasters make the design of the humanitarian supply chain network (HSCN) a crucial strategic issue in a pre-disaster scenario. The HSCN design problem deals with the location/allocation of emergency response facilities (ERFs). This paper aims to propose and demonstrate how to design an efficient HSCN configuration under the risk of ERF disruptions.

This paper considers four performance measures simultaneously for the HSCN design by formulating a weighted goal programming (WGP) model. Solving the WGP model with different weight values assigned to each performance measure generates various HSCN configurations. This paper transforms a single-stage network into a general two-stage network, treating each HSCN configuration as a decision-making unit with two inputs and two outputs. Then a two-stage network data envelopment analysis (DEA) approach is applied to evaluate the HSCN schemes for consistently identifying the most efficient network configurations.

Among various network configurations generated by the WGP, the single-stage DEA model does not consistently identify the top-ranked HSCN schemes. In contrast, the proposed transformation approach identifies efficient HSCN configurations more consistently than the single-stage DEA model. A case study demonstrates that the proposed transformation method could provide a more robust and consistent evaluation for designing efficient HSCN systems. The proposed approach can be an essential tool for federal and local disaster response officials to plan a strategic design of HSCN.

This study presents how to transform a single-stage process into a two-stage network process to apply the general two-stage network DEA model for evaluating various HSCN configurations. The proposed transformation procedure could be extended for designing some supply chain systems with conflicting performance metrics more effectively and efficiently.

Finding efficient disaster recovery center location-allocation-routing (DRCLAR) network schemes play a vital role in the disaster recovery logistics network (DRLN) design. The purpose of this paper is to propose and demonstrate how to design efficient DRCLAR network schemes under the risk of facility disruptions as a part of the disaster relief activities.

A goal programming (GP) model is formulated to consider four performance measures simultaneously for the DRCLAR design. The cross-evaluation based-super efficiency data envelopment analysis (DEA) approach is applied to better evaluate the DRCLAR network schemes generated by solving the GP model so that more efficient network schemes can be identified.

The proposed approach identifies more efficient DRCLAR network schemes consistently among various network schemes generated by GP. We find that combining these two methods compensates for each method's weaknesses and enhances the discriminating power of the DEA method for effectively identifying and ranking the network schemes.

This study presents how to generate balanced DRCLAR network schemes and how to evaluate various network schemes for identifying efficient ones. The proposed procedure of developing and evaluating them could be extended for designing some disaster recovery/relief supply chain systems with conflicting performance measures.

Emergency relief supply chain (ERSC) design is an important strategic decision that significantly affects the overall performance of emergency management activities. The performance of an ERSC can be measured by several performance measures some of which may conflict with each other. The purpose of this paper is to propose an ERSC design framework by simultaneously taking total logistics cost (TLC), risk level, and amount of demands covered in an ERSC into consideration.

The study considers TLC of an ERSC as the sum of logistics cost from distribution warehouses (DWHs) to Break of Bulbs (BOBs) and from BOBs to affected neighborhoods. The risk level of an ERSC is measured by estimating the expected number of disrupted relief items (EDI) distributed from DWHs through BOBs to neighborhoods. The covered demand (CDM) is defined as total populations that are supported in case of an emergency, the populations within the maximal coverage distance (MCD) from relief facilities. Based on these performance measures, the authors formulate a Goal Programming (GP) model to distribute emergency relief items to affected locations. Ideal values of these performance measures are decided, and the GP model seeks to minimize the weighted sum of the percentage deviations of those performance measures from the ideal values. The relationships among performance measures have been thoroughly analyzed through detailed trade-off studies under two realistic case studies by changing weights of each performance measure.

Three performance measures are interdependent over specific values of weights. TLC and EDI have a trade-off relationship when the weight on each measure increases. TLC and CDM also have a trade-off relationship when the weight on EDI increases. However, this relationship becomes less apparent when the MCD increases. EDI and CDM also have the same trade-off relationship when the weight on TLC changes. Therefore, decision makers should thoroughly analyze these trade-off relationships when they design ERSCs. Overall, the study identified that an ERSC with higher MCD outperforms one with lower MCD in terms of TLC, EDI, and CDM.

The study presents a design framework to generate more balanced ERSCs by simultaneously taking three conflicting performance measures into consideration, and demonstrated the feasibility of the framework through realistic case studies. The trade-off analysis provides useful insights and theoretical knowledge to researchers and practitioners in the discipline of emergency logistics management. The results from this study are expected to contribute to the development of more balanced ERSCs.