SCRO / Journées de l'optimisation

HEC Montréal, 29-31 mai 2023


HEC Montréal, 29 — 31 mai 2023

Horaire Auteurs Mon horaire

DMI Disruption Management I

31 mai 2023 10h30 – 12h10

Salle: Sony (jaune)

Présidée par Mateo Espitia Ibarra

4 présentations

  • 10h30 - 10h55

    Handling Ambiguity in the Humanitarian Supply Chain Network Design Problem

    • Mohammad Daneshvar, prés., ESG-UQAM
    • Sanjay Dominik Jena, Université du Québec à Montréal
    • Walter Rei, ESG-UQAM

    After a natural disaster, both the state of the impacted region and the severity of the crisis among the affected population are uncertain. To establish an effective supply chain network to support humanitarian activities, decision-makers first estimate the uncertain parameters using multiple data sources, which may include surveys, satellite imagery, governmental reports, and media. However, estimates may turn out to be quite different from one another, leading to ambiguity in the informational context in which the planning process of humanitarian relief operations occurs.

    In this presentation, I will present our ongoing work on handling inconsistent estimates stemming from the use of multiple data sources when modeling and solving a Humanitarian Supply Chain Network design problem post-disaster. Specifically, we study the problem of designing the distribution network to perform the storage and distribution of critical supplies to the affected population over a given time horizon. Multiple optimization models under uncertainty are developed to formulate the problem, while explicitly considering the uncertainty that is related to the population needs and the network's transportation and storage capacities. The models are then used to solve a series of instances based on data from the 2018 earthquake in Indonesia to study the impacts that different ambiguity patterns have on the results obtained using the different proposed models.

  • 10h55 - 11h20

    Dynamic Supply Chain Planning under Uncertainty in Disaster Relief Operations

    • Kai Huang, prés., McMaster University
    • afshin kamyabniya, McMaster University

    This paper focuses on the response to a Disaster Relief Operation (DRO) involving people with varying types of demands and multiple potential destinations in a dynamic supply chain network setting. First, we look to determine the optimal destination for each people at the DRO. Second, we look to determine the timing and size of the capacity of supply increases at the various supply points in the network. Such a problem is modelled as a sequential stochastic decision-making problem. To solve and evaluate the model for large-scale instances, a two-step column generation algorithm is developed to optimize the delivery service of supply points under various sources of uncertainty.
    The model is then applied to a real DRO case study. The results demonstrate that the performance of the proposed model and the solution methodology can provide timely policies for people's transportation and the capacity planning problem that ensures timely access to relevant services as well as a robust capacity plan. It would be beneficial for relief response managers to apply the proposed model and its derived policies compared with currently implemented policies.

  • 11h20 - 11h45

    An Overland Search and Rescue Problem With Search Area Selection

    • Saeid Abbasiparizi, prés., Ph.D. Student at the Department of Operations and Decision Systems, Université Laval, Québec, Canada
    • Michael Morin, Assistant Professor at the Department of Operations and Decision Systems, Université Laval, Québec, Canada
    • Irène Abi-Zeid, Full Professor at the Department of Operations and Decision Systems, Université Laval, Québec, Canada

    Many Search and rescue incidents occur in Canada every year and, some will require the deployment of major search operations. When search operations are initiated, tasking search and rescue units, such as helicopters, in a timely and efficient fashion is of the essence. In this project, we propose to model the problem of selecting multiple search areas along with the order in which they are visited in the context of overland search operations by aircraft. Our mixed-integer programming model maximizes the probability of success of the operation (finding the search objects) over a set of disjoint rectangular areas under available effort and operational constraints. We use our model to compute the optimal set of areas (rectangles) to be searched along with the shortest flight plan to search them. Although models for search operations planning exist in the literature, our approach integrates the search area selection and the planning phases. We present results on the scalability of several approaches including the complete mixed-integer program and decomposition approaches such as the Benders decomposition algorithm.

    Keywords: Search theory, Search and rescue, Search area selection, Mixed-integer programming, Benders decomposition algorithm.

  • 11h45 - 12h10

    An integrated approach to debris recycling operations

    • Mateo Espitia-Ibarra, prés., ESG - UQAM
    • Ana María Anaya Arenas, ESG - UQAM
    • Hani Zbib, ESG - UQAM

    In a disaster’s aftermath, the presence of large debris quantities hampers response and relief operations, and can have a significant impact on the economy and well-being of affected communities. Debris management operations is one of the most costly, complicated, and time-consuming post-disaster activities. However, with an increasing global focus on sustainability, it is important to move from approaches that efficiently collect and then dispose of debris, to approaches that efficiently collect and then recycle debris.
    This talk presents a novel approach that aims to minimize the economic, environmental, and psychological impacts of debris recycling operations. The problem formulation incorporates decisions related to sorting, compacting, and transporting debris from the demand points (DP) to their final recycling treatment facilities, passing through temporary debris management sites (TDMS). The problem therefore considers what debris sorting and compaction technologies to adopt, whether to deploy them on-site at the DP or off-site at the TDMS, the location and capacity allocation of the TDMS, and routing vehicles in the network. Implementing this approach presents a practical and efficient solution to the challenges of managing post-disaster debris recycling, since it enables communities to sustainably manage the debris generated by natural disasters and hence mitigate their impacts.