•Studies on supply chain network design under uncertainty are reviewed.•Uncertain decision-making environments and uncertainty sources are categorized.•The studies are investigated in terms of supply ...chain management and optimization aspects.•Literature's gap and a list of future research directions are highlighted.
Supply chain network design (SCND) is one of the most crucial planning problems in supply chain management (SCM). Nowadays, design decisions should be viable enough to function well under complex and uncertain business environments for many years or decades. Therefore, it is essential to make these decisions in the presence of uncertainty, as over the last two decades, a large number of relevant publications have emphasized its importance. The aim of this paper is to provide a comprehensive review of studies in the fields of SCND and reverse logistics network design under uncertainty. The paper is organized in two main parts to investigate the basic features of these studies. In the first part, planning decisions, network structure, paradigms and aspects related to SCM are discussed. In the second part, existing optimization techniques for dealing with uncertainty such as recourse-based stochastic programming, risk-averse stochastic programming, robust optimization, and fuzzy mathematical programming are explored in terms of mathematical modeling and solution approaches. Finally, the drawbacks and missing aspects of the related literature are highlighted and a list of potential issues for future research directions is recommended.
We address a multi-period supply chain (SC) network design where demands of customers depend on facilities serving them based on their delivery lead-times. Potential customer demands are stochastic, ...and facilities’ capacity varies randomly because of possible disruptions. Accordingly, we develop a multi-stage stochastic program, and model disruptions’ effect on facilities’ capacity. The SC responsiveness risk is limited and, to obtain a resilient network, both mitigation and contingency strategies are exploited. Computational results on a real-life case study and randomly generated problem instances demonstrate the model’s applicability, risk-measurement policies’ performance, and the influence of mitigation and contingency strategies on SC’s resiliency.
•Flexibility in CLSCN design to cover demands and returns based on market conditions.•A novel hybrid robust-stochastic programming approach.•Scenario generation using Latin Hypercube Sampling with ...backward reduction.•Accelerated Benders decomposition with valid inequalities and Pareto-optimal cuts.
Environmental, social and economic concerns motivate the operation of closed-loop supply chain networks (CLSCN) in many industries. We propose a novel profit maximization model for CLSCN design as a mixed-integer linear program in which there is flexibility in covering the proportions of demand satisfied and returns collected based on the firm's policies. Our major contribution is to develop a novel hybrid robust-stochastic programming (HRSP) approach to simultaneously model two different types of uncertainties by including stochastic scenarios for transportation costs and polyhedral uncertainty sets for demands and returns. Transportation cost scenarios are generated using a Latin Hypercube Sampling method and scenario reduction is applied to consolidate them. An accelerated stochastic Benders decomposition algorithm is proposed for solving this model. To speed up the convergence of this algorithm, valid inequalities are introduced to improve the lower bound quality, and also a Pareto-optimal cut generation scheme is used to strengthen the Benders optimality cuts. Numerical studies are performed to verify our mathematical formulation and also demonstrate the benefits of the HRSP approach. The performance improvements achieved by the valid inequalities and Pareto-optimal cuts are demonstrated in randomly generated instances.
In this paper, we address a multi-period supply chain network redesign problem in which customer zones have price-dependent stochastic demand for multiple products. A novel multi-stage stochastic ...program is proposed to simultaneously make tactical decisions including products’ prices and strategic redesign decisions. Existing uncertainty in potential demands of customer zones is modeled through a finite set of scenarios, described in the form of a scenario tree. The scenarios are generated using a Latin Hypercube Sampling method and then a forward scenario construction technique is employed to create a suitable scenario tree. The multi-stage stochastic problem is formulated as a mixed-integer linear programming model and then Benders decomposition algorithm is applied for solving it. Numerical results demonstrate the significance of the stochastic model as well as the good performance of Benders algorithm. The scenario tree construction method is also evaluated in terms of out-of-sample and in-sample stability. Finally, several key managerial and practical insights in terms of pricing issues are highlighted.
•We address integrated resource sharing and demand redistribution during a pandemic.•We propose a multi-stage stochastic program under resources’ demand uncertainty.•We propose a data-driven ...decision-making framework.•We investigate two real-life case studies for the COVID-19 pandemic.
We study resource planning strategies, including the integrated healthcare resources’ allocation and sharing as well as patients’ transfer, to improve the response of health systems to massive increases in demand during epidemics and pandemics. Our study considers various types of patients and resources to provide access to patient care with minimum capacity extension. Adding new resources takes time that most patients don't have during pandemics. The number of patients requiring scarce healthcare resources is uncertain and dependent on the speed of the pandemic's transmission through a region. We develop a multi-stage stochastic program to optimize various strategies for planning limited and necessary healthcare resources. We simulate uncertain parameters by deploying an agent-based continuous-time stochastic model, and then capture the uncertainty by a forward scenario tree construction approach. Finally, we propose a data-driven rolling horizon procedure to facilitate decision-making in real-time, which mitigates some critical limitations of stochastic programming approaches and makes the resulting strategies implementable in practice. We use two different case studies related to COVID-19 to examine our optimization and simulation tools by extensive computational results. The results highlight these strategies can significantly improve patient access to care during pandemics; their significance will vary under different situations. Our methodology is not limited to the presented setting and can be employed in other service industries where urgent access matters.
•We present a mixed integer programming model for healthcare districting problem.•We extend the model for the robust decision-making under uncertainty.•We propose novel graph-based GA algorithms for ...solving the problem.•A real-life case study is investigated.
In this paper, we address a districting problem motivated by a real-world case study to partition residential areas of a region for a healthcare system operation under interval demand uncertainty of health services. A mixed-integer programming model is presented based on the graph theory that enforces contiguity constraints on districts in addition to other corresponding practical criteria. Furthermore, robust optimization approaches are extended to address the existing uncertainty. To deal with the problem’s computational intractability, an improved genetic algorithm is developed in accordance with the graph-based nature of the problem obtaining near-optimal solutions on large-sized instances. Extensive computational results are presented on a real-world case study and several randomly generated instances to evaluate the applicability of the models, performance of the presented robustness measure, and effectiveness of the solution approach. Furthermore, a hierarchical districting approach that enables decision makers to obtain districting decisions in various levels of health services is examined. Sensitivity analyses on main parameters are performed to derive some managerial insights that can help practitioners in providing suitable and homogeneous health services in a geographical area.
We study a coordinated clinic and surgery appointment scheduling problem for in‐advance scheduling of surgical patients. Our models seek to provide timely access to care by coordinating clinic and ...surgery appointments to ensure that patients can see a surgeon in the clinic and (if needed) schedule their surgery within a maximum wait time target based on patient classes. There are different types of uncertainty including the number of appointment requests, whether a patient requires surgery, and surgery durations. We develop an integrated multi‐stage stochastic and distributionally robust optimization (IMSDRO) approach to determine the optimal clinic and surgery dates for patients such that the access target constraints are satisfied, and the clinical and surgical overtimes are minimized. The IMSDRO approach synergizes multi‐stage stochastic optimization with distributionally robust optimization to simultaneously incorporate multiple types of uncertainties by including stochastic scenarios for appointment request arrivals and ambiguity sets for surgery durations. Several new transformations are introduced to turn the nonlinear model derived from the IMSDRO approach to a tractable one, and a constraint generation algorithm is developed to solve it efficiently. We propose a data‐driven rolling horizon procedure to facilitate implementation. We use case data to assess the performance of our policies. The results suggest that our policy can significantly improve surgical access delay times compared to the current practice. Our methodology is not limited to a particular setting and can be applied to other service industries where access delay matters.
This dissertation is broadly about developing new personalized data-driven learning and optimization methods with theoretical performance guarantees for three important applications in healthcare ...operations management and medical decision-making. In these research problems, we are dealing with longitudinal settings, where the decision-maker needs to make multi-stage personalized decisions while collecting data in-between stages. In each stage, the decision-maker incorporates the newly observed data in order to update his current system's model or belief, thereby making better decisions next. This new class of data-driven learning and optimization methods indeed learns from data over time so as to make efficient and effective decisions for each individual in real-time under dynamic, uncertain environments. The theoretical contributions lie in the design and analysis of these new predictive and prescriptive learning and optimization methods and proving theoretical performance guarantees for them. The practical contributions are to apply these methods to resolve unmet real-world needs in healthcare operations management and medical decision-making so as to yield managerial and practical insights and new functionality.
Environmental, social and economic concerns motivate the operation of closed- loop supply chain networks (CLSCN) in many industries. We propose a novel profit maximization model for CLSCN design as a ...mixed-integer linear program in which there is flexibility in covering the proportions of demand satisfied and returns collected based on the firm's policies. Our major contribution is to develop a novel hybrid robust-stochastic programming (HRSP) approach to simultaneously model two different types of uncertainties by including stochastic scenarios for transportation costs and polyhedral uncertainty sets for demands and returns. Transportation cost scenarios are generated using a Latin Hypercube Sampling method and scenario reduction is applied to consolidate them. An accelerated stochastic Benders decomposition algorithm is proposed for solving this model. To speed up the convergence of this algorithm, valid inequalities are introduced to improve the quality of lower bound, and also a Pareto-optimal cut generation scheme is used to strengthen the Benders optimality cuts. Numerical studies are performed to verify our mathematical formulation and also demonstrate the benefits of the HRSP approach. The performance improvements achieved by the valid inequalities and Pareto-optimal cuts are demonstrated in randomly generated instances.