Management Science and Engineering

Permanent URI for this collectionhttps://uwspace.uwaterloo.ca/handle/10012/9910

This is the collection for the University of Waterloo's Department of Management Science and Engineering.

Research outputs are organized by type (eg. Master Thesis, Article, Conference Paper).

Waterloo faculty, students, and staff can contact us or visit the UWSpace guide to learn more about depositing their research.

Browse

Now showing 1 - 7 of 7

Distribution Planning with Consolidation - A Two-Stage Stochastic Programming Approach
(University of Waterloo, 2017-09-07) Alnaggar, Aliaa; Gzara, Fatma; Bookbinder, James H.
The distribution planning problem with consolidation center(s) addresses the coordination of distribution activities between a set of suppliers and a set of customers, through the use of intermediate facilities in order to achieve savings in transportation cost. We study the problem from the perspective of a third-party logistics provider (3PL) that is coordinating shipments between suppliers and customers. Given customer demand of products from different suppliers, the goal is to consolidate the shipments in fewer high volume loads, from suppliers to the consolidation center(s) and from the consolidation center(s) to customer. We assume that suppliers have a finite set of transportation options, each with a given capacity and time of arrival at the consolidation center(s). Similarly, customers have a set of transportation options, each with a given capacity and dispatch time from the consolidation center(s). The 3PL wants to determine the optimal transportation options, or shipment schedule, and the allocation of shipments to transportation options from suppliers to consolidation center(s), and from consolidation center(s) to customers, that minimize the total transportation cost and holding cost at the consolidation center. The literature studies many variations of this problem, which assume deterministic demand. This thesis extends the problem for stochastic demand and formulates it as a two-stage stochastic programming model. We model the case where the choice of transportation options is a \textit{contractual} decision, and a 3PL needs to decide on which options to reserve for a given planning period subject to stochastic customer demand. Therefore, the choices of transportation options are the stage one variables in the two-stage stochastic program. The second stage variables, which are decisions that are made after the uncertainty conditions become known, represent the allocation of orders to reserved transportation options as well as shipping orders through a spot-market carrier, at a greater transportation cost. Because of the high computational demand of the model, the integer L-shaped method is applied to decompose the problem. To increase the efficiency of the algorithm, we experiment with three valid cuts with the goal of generating stronger cuts than the L-cut. We also apply three algorithm enhancement techniques to speed up the convergence of the algorithm. Numerical results show that the performance of our proposed methodology and valid cuts is comparable to that of CPLEX. We suggest promising areas for future work to further improve the computational efficiency of our decomposition algorithm.
A Dual Toll Policy for Regulating the Transportation of Hazardous Materials
(University of Waterloo, 2018-08-03) Zhang, Huiwen; Bookbinder, James H.; Ke, Yi (Ginger)
The transportation of hazardous materials (hazmat) has drawn significant attention from various stakeholders due to the undesirable impacts on the environment and public health. Focusing on the connection between the traffic and the risk associated with the hazmat shipments, the present research aims to assist the regulator in designing a policy of dual tolls, imposed on both hazmat and non-hazmat shipments, to mitigate the hazmat risk in a road network. A bi-objective bi-level programming formulation is constructed. To be specific, the upper-level model indicates the regulator’s decision problem, minimizing the maximum link risk and the total network risk by imposing a dual-toll policy on any carrier. The lower level jointly considers the decisions of multiple hazmat carriers and non-hazmat travelers, minimizing the total transportation cost, including the toll cost. (By “non-hazmat traveler", we mean both people who carry and do not carry products.) Given the bi-level structure and the non-linear nature, a solution procedure with two parts is designed. First, we develop two alternative linearization approaches. One is piecewise linearization, transforming the non-linear terms into linear ones. The other applies the Frank-Wolfe algorithm, an iterative first-order optimization algorithm. Then a genetic-algorithm-based methodology will integrate both levels. Computational experiments on different sizes of networks are performed to demonstrate the effectiveness of the model. Various analyses, involving trade-offs, sensitivities, and examination of convergence, are conducted to provide additional managerial insights. These can be used to facilitate stakeholders’ decision making.
Location-Routing Problems with Economies of Scale
(University of Waterloo, 2016-10-26) Pi, Xiaoyang; Bookbinder, James H.
The purpose of the location-routing problem is to select facility locations, assign customers to facilities, and design routes between facilities and customers. The most common objective is to achieve minimal cost. In general, for the standard location-routing problem, the total cost includes xed costs to open Distribution Centers (DCs) plus the transportation costs from facilities to customers. In this thesis, we also consider the variable cost of facilities' operations. Two forms of the variable cost are proposed. One is a linear cost function with a constant variable (operating) cost per unit; the other employs a concave function of total throughput at any DC as its cost model. The latter is studied because economies of scale can be achieved for large facilities. By economies of scale, we mean that the variable cost per unit is a decreasing function of the number of units of throughput; that relationship is expressed by a concave function. Two solution methods are developed, both based on a genetic algorithm. After some preliminary tests, one approach is employed for further testing. Computational experiments of the model without variable cost are performed on published data sets. Then extensive testing is done on modi ed data sets for cases with operating cost but without economies of scale, and for other cases when economies of scale are present. Analysis of the in influence of economies of scale is provided. First, we brie y test how parameter values a ect the economies of scale. Then we extensively analyze the tradeo s between operating costs of facilities and transportation costs. Conclusions are drawn and further research is suggested.
Lot-Sizing of Several Multi-Product Families
(University of Waterloo, 2018-04-19) Bayley, Tiffany; Bookbinder, James H.
Production planning problems and its variants are widely studied in operations management and optimization literature. One variation that has not garnered much attention is the presence of multiple production families in a coordinated and capacitated lot-sizing setting. While its single-family counterpart has been the subject of many advances in formulations and solution techniques, the latest published research on multiple family problems was over 25 years ago (Erenguc and Mercan, 1990; Mercan and Erenguc, 1993). Chapter 2 begins with a new formulation for this coordinated capacitated lot-sizing problem for multiple product families where demand is deterministic and time-varying. The problem considers setup and holding costs, where capacity constraints limit the number of individual item and family setup times and the amount of production in each period. We use a facility location reformulation to strengthen the lower bound of our demand-relaxed model. In addition, we combine Benders decomposition with an evolutionary algorithm to improve upper bounds on optimal solutions. To assess the performance of our approach, single-family problems are solved and results are compared to those produced by state-of-the-art heuristics by de Araujo et al. (2015) and Süral et al. (2009). For the multi-family setting, we first create a standard test bed of problems, then measure the performance of our heuristic against the SDW heuristic of Süral et al. (2009), as well as a Lagrangian approach. We show that our Benders approach combined with an evolutionary algorithm consistently achieves better bounds, reducing the duality gap compared to other single-family methods studied in the literature. Lot-sizing problems also exist within a vendor-managed-inventory setting, with production-planning, distribution and vehicle routing problems all solved simultaneously. By considering these decisions together, companies achieve reduced inventory and transportation costs compared to when these decisions are made sequentially. We present in Chapter 3 a branch-and-cut algorithm to tackle a production-routing problem (PRP) consisting of multiple products and customers served by a heterogeneous fleet of vehicles. To accelerate the performance of this algorithm, we also construct an upper bounding heuristic that quickly solves production-distribution and routing subproblems, providing a warm-start for the branch-and-cut procedure. In four scenarios, we vary the degree of flexibility in demand and transportation by considering split deliveries and backorders, two settings that are not commonly studied in the literature. We confirm that our upper bounding procedure generates high quality solutions at the root node for reasonably-sized problem instances; as time horizons grow longer, solution quality degrades slightly. Overall costs are roughly the same in these scenarios, though cost proportions vary. When backorders are not allowed (Scenarios 1 and 3), inventory holding costs account for over 90% of total costs and transportation costs contribute less than 0.01%. When backorders are allowed (Scenarios 2 and 4), most of the cost burden is shouldered by production, with transportation inching closer to 0.1% of total costs. In our fifth scenario for the PRP with multiple product families, we employ a decomposition heuristic for determining dedicated routes for distribution. Customers are clustered through k-means++ and a location-alloction subproblem based on their contribution to overall demand, and these clusters remain fixed over the entire planning horizon. A routing subproblem dictates the order in which to visit customers in each period, and we allow backorders in the production-distribution routine. While the branch-and-cut algorithm for Scenarios 1 through 4 quickly finds high quality solutions at the root node, Scenario 5's dedicated routes heuristic boasts high vehicle utilization and comparable overall costs with minimal computational effort.
Optimization and Comparison of Manual and Semi-Automated Material Handling in a Cross-Dock Using Discrete-Event Simulation
(University of Waterloo, 2018-08-21) Natarajan, Saravanan; Bookbinder, James H.
A Cross-Dock (CD) is a synchronized unit of a supply chain network, used to sort the goods received from inbound trucks (from a warehouse or factory), and load those products to outbound trucks (for delivery of the goods to retail stores in the supply chain network). Most cross-docks use forklifts, and other manual material handling equipment (MHE) to process the goods on pallets received from inbound trucks. Those pallets are sorted and loaded onto outbound trucks. With the advancements in robotics, it could be bene cial to employ semi-automated material handling techniques in a CD, rather than solely relying on manual material handling. In this thesis, the scope of self-driving vehicles (SDVs) in one such semi-automated cross-dock facility is studied. We compare the cases of purely manual and semi-automated material handling in a cross-dock. Using simulation, we modelled two cross-dock facilities, one with forklifts only and with a mixture of forklifts and SDVs. Simulation was thus employed to mimic the CD's material handling process, to compare the two MHE con gurations. Then the built cross-dock simulation models were optimized using the response surface methodology and mixed integer non-linear programming (MINLP), to achieve the optimal MHE con guration for those facilities operating with the desired levels of performance metrics. Thereby the manual and semi-automated cross-dock with similar performance (and optimal MHE con gurations) are compared and the scope of SDVs in a cross-dock is evaluated. Conclusions are given, and opportunities for further research are presented.
Optimization Models for the Perishable Inventory Routing Problem
(University of Waterloo, 2019-08-29) Singla, Ajitesh Rajiv; Bookbinder, James H.; Naoum-Sawaya, Joe
In this thesis, three models for the Perishable Inventory Routing Problem (PIRP) are explored. The first model of the inventory routing problem considers one perishable product and known (deterministic) demands in the context of consignment inventory. The objective of the problem maximizes the profit of the supplier who owns all the inventory until it is sold. The supplier gives a fixed percentage discount on the selling price of the product as it deteriorates. The shelf life and the number of vehicles used in this problem is also fixed. Computational results, comparing the PIRP with the standard Inventory Routing Problem (IRP), are presented. Based on the calculations using the branch-and-cut algorithm, we conclude that adding perishability to the IRP leads to better inventory management as a result of which fresher products are delivered to the customers. The PIRP model is also solved using the Benders Decomposition algorithm. The results indicate that in the above stated problem, the branch-and-cut algorithm yields better results as compared to the Benders Decomposition algorithm. The second model extends the PIRP to consider uncertain (stochastic) demands (SPIRP). This is done to make the model comparable to a real-life problem. Several demand scenarios are created within a fixed range to capture the uncertainty with respect to the perishable product. Since estimating the actual probability is difficult, a large number of equally probable demand scenarios are assumed. The model was run on the branch-and-cut and the Benders Decomposition algorithm. The results generated were then compared. It was inferred that even when demand scenarios enable easy decomposition of the problem, the Benders algorithm performs worse than the branch-and-cut algorithm. In the third model, the robust formulation of the SPIRP is proposed to resolve the above-mentioned limitations. Robustness was added to the SPIRP to enable the use of a small number of scenarios to obtain solutions that are competitive with modeling a large number of scenarios in the case of SPIRP. An innovative way of formulating the robust counterpart of the SPIRP was developed while keeping the probability of each demand scenario uncertain. An algorithm was devised to compare the effectiveness of the robust model to the deterministic and stochastic models. Computational results compare the average profit values generated by the three models. It is concluded that while the deterministic model captures no uncertainty, the stochastic model with many scenarios accounts for the most demand uncertainty; the robust model, through the use of far fewer scenarios, can account for a significant uncertainty in demand. Another interpretation of the results is that an increased number of robust scenarios has a significant effect on the average profit values of that model.
Shipment Scheduling In Hub Location Problems
(University of Waterloo, 2017-07-26) Masaeli, Mobina; Bookbinder, James H.; Alumur Alev, Sibel
In this thesis, we incorporate shipment scheduling decisions into hub location problems. Our aim is to determine optimal locations of hubs, hub network structure, and the number of vehicles to operate on the hub network as well as the time period of dispatching each vehicle from a hub. We develop mathematical models for di erent versions of this problem. We initially propose a mixed-integer shipment scheduling hub network design model where the costs of holding freight are negligible. We then expand the model to keep track of the holding decisions where the holding costs are not negligible. We further analyze the shipment scheduling model with holding costs when di erent types of vehicles are available to operate on the inter-hub links. We investigate the impact of shipment scheduling decisions and holding costs on hub network con gurations, routing decisions, and total cost of the network. We solve the models on instances from a new USAF dataset with real data.

Browse

Browsing Management Science and Engineering by Author "Bookbinder, James H."