Applications of Multi-Criteria and Game Theory Approaches

Vehicle routing problems play a central role in logistics. These combinatorial optimization problems have attracted more and more attention these last five decades both in theory and in practice. However, main contributions are dedicated to the single criterion optimization problems. The goal of this chapter is to provide the recent key references dedicated to multi-criteria studies in transportation logistics and especially on vehicle routing problems and to present some interesting research directions.

This chapter deals with the use of multi-objective approaches in the field of assembly line design. The design of assembly or transfer lines is a very important industrial problem, which involves various difficult and interconnected optimization problems. A review of the main multi-objective optimization methods used for these problems is presented and discussed. A case study is also described in order to highlight some interesting properties associated with such multi-objective problems.

Determining an appropriate storage policy is a critical issue in warehouse management. Storage policies address location assignment of stock keeping units (SKUs) in warehouses. An effective storage policy should not only provide the minimization of transportation and inventory costs, but also increase the level of service available to the internal and external customers. When selecting a storage policy, parameters cannot be frequently determined as crisp values. Fuzzy logic is utilized in many engineering applications so as to handle imprecise data. Moreover, information axiom, the second axiom of axiomatic design (AD), performs the selection of the alternative that mostly satisfies the functional requirements of decision makers. This chapter provides a fuzzy information axiom basis for storage policy selection. After providing background information about storage policies as well as storage assignment models, a fuzzy information axiom-oriented model is introduced. Then, the decision-making model is validated by an application in a company from automotive industry.

Technological innovations and changing customer trends brought by globalization has led tough competition among various industries throughout the globe. Their assiduous efforts to develop new product is crucial for survival. To overcome this problem and to develop a quality product that generates revenue, a dynamical multi-objective evolutionary algorithm(DMOEA) incorporated with quality function deployment (QFD) and fuzzy analytic network process (FANP) is proposed. The proposed approach considers goals such as new product development (NPD) time and cost, technological advancement, and manufacturability for selection of the most suitable product technical requirements (PTRs). A case study of software development is included to demonstrate the effectiveness of the proposed approach and the obtained results are discussed.

The river and maritime transport represents an attractive alternative to land and air transport. The containerization allows the industries to save costs thanks to the standardization of dimensions. The container terminal has to manage container traffic at the crossroads of land road and railway. In this chapter, we propose to optimize, simultaneously, the storage problem and the quayside transport problem. In a space storage, we have several blocks and each one has its storage cost. The first aim is to minimize the cost storage of containers. These latter are loaded into vessels, the vehicles have to transport the containers from blocks to quays (of vessels). Thus, the second aim consists to minimize the distance between the space storage and the quays. The optimization methods of operations research in container terminal operation have become more and more important in recent years. Objective methods are necessary to support decisions. To solve this multi-objective problem, we develop two resolution methods based on metaheuristic approach called ant colony algorithm. The first one is multi-objective ant colony optimization (noted MOACO) and the second one is the MOACO with a local search (called MOACO-LS), good promising results are given.

Mobile supply chain management can help manufacturers to reduce cost and improve supply chain performances. However, the decisions to adopt mobile supply chain management are complex as it involved multi-criterion decisions that need to be considered by manufacturing firms. This research aims to predict the factors that can lead to successful mobile supply chain management adoption. Variables from the technology-organization-environment (TOE) model were used as predictors for this research. A non-compensatory adoption decision process is modeled using neural network analysis. Data was collected from 192 manufacturing firms. Our results showed that some of the strongest predictors for mobile supply chain management adoption are senior management support, security perceptions, technology integrations, and financial and technical competence. Firm size and environmental factors on the other hand have less predictive power than technological and organizational factors on mobile supply chain management adoption decisions.

This chapter deals with the performance expression problematic in an industrial continuous improvement process. Performance expressions are the purpose of performance indicators and performance measurement systems (PMSs). We focus particularly on the elementary aspect of such an expression. The elementary performance expression is the constitutive element of the PMSs, being defined through the achievement degree of a considered objective, while other types of expressions are involved in PMSs, with regard to the multi-criteria and multilevel aspects of the objectives. The computation of the objective achievement brings together the objective declaration, the acquired measurement that reflects the reached state and the comparison of these parameters. By revisiting previous works handled in this field, we consider that elementary performance expression is modelled by a mathematical function that compares the objective to the measurement. Conventional Taylorian ratio and difference are highlighted. The qualitative or quantitative characteristic of the data, the flexibility concerning the objective declaration and the measurements errors lead us to use the fuzzy subset theory as a unified framework for expressing performance. It also leads to new approaches which are beyond comparison functions.

Measurement of supply chain performance is an important issue to identify success, to understand processes, to figure out problems and where improvements are possible as well as provide facts for decision-making. Using classical performance measurement techniques, it may not be possible to incorporate judgments of decision makers comprehensively. Hence, we propose a fuzzy multi-criteria evaluation method for this purpose in the framework of supply chain performance measurement. Fuzzy DEMATEL is used to prioritize the performance measurement criteria of supply chain. We also present a sensitivity analysis using different linguistic scales.

This chapter focuses on selecting the route in international intermodal freight transportation network considering the following characteristics, first and foremost multi-objective: minimization of travel time and travel cost, later schedules and delivery times of every service provider in each pair of location, and lastly variable cost must be included in every location. The study aims to formulate the problem into mixed integer linear programming (MILP) model and develop an algorithm which encompassing all the above essential characteristics. It is NP-hard problem; it follows the proposed algorithm (nested partitions method) that is heuristic and multi-attribute decision-making (MADM) method. An illustrative experiment is considered and our proposed algorithm is applied to obtain an effective and efficient solution.

This chapter discusses the issue of integrating inventory and distribution optimization together with game theory to effectively manage supply networks. Inventory and distribution simultaneously optimization is a challenging problem aiming at coordinating decisions related to inventory management with those related to transportation scheduling. This problem is known as the inventory routing problem (IRP) and is an underlying optimization model for supply networks implementing a vendor managed inventory (VMI) strategy. Game theory, and in particular cooperative games, involves several decision-makers willing to coordinate their strategies and share the payoff. In particular, coalitions of decision-makers can make binding agreements about joint strategies, pool their individual payoffs, and redistribute the total in some specified way. In a supply and distribution context, the manager of a franchising business must decide how much inventory to carry. Naturally, the manager of each sales-points wishes to carry an as low as possible amount of inventory and at the same time have enough inventory to cover all demand and not miss any potential sale. One possibility to achieve these two contrasting goals is to allow cooperation among the sales-points and trade the product at some fair price. Sales-points with an excess inventory may want to sell that surplus to other sales-points in the same cluster or coalition, facing a larger than expected demand. The game consists in determining clusters of sales-points which are willing to cooperate, a fair trade-price, and inventory quantities to be carried by each sales-points to minimize the total costs and maximize the total sales. In other models, the total cost of transportation between a depot and a set of customers must be divided among them and the game considers the synergies in the determination of the individual costs.

In this chapter, we consider the problem of determining an optimal set of winning suppliers in a procurement auction where the buyer wishes to procure high volumes of a homogeneous item in a staggered way in accordance with a predefined schedule and the suppliers respond with bids that specify volume discounts and also delivery lead times. We show that the winner determination problem, which turns out to be a multi-objective optimization problem, cannot be satisfactorily solved by traditional methods of multi-objective optimization. We formulate the problem first as an integer program with constraints capturing lead time requirements and show that the integer program is an extended version of the multiple knapsack problems. We discover certain properties of this integer program and exploit the properties to simplify it to a 0–1 mixed integer program (MIP), which can be solved more efficiently. We next explore a more efficient approach to solving the problem using a linear relaxation of the 0–1 MIP in conjunction with a greedy heuristic. Using extensive numerical experimentation, we show the efficacy of the 0–1 MIP and the proposed heuristic.

This chapter analyzes the process of forming a coalition within a corporate network. The objective of the partner companies is to create a multistage manufacturing system, which generates a chain of increased value from raw materials to end-user market. This process is studied by cooperative game theory, through the key problems of maximizing the total profit and distributing it among the members of the coalition. To construct a pay-off policy that is both stable and fair, the study proposes to represent the productive resources of the firms not only by their capacity, but also by the work in progress (WIP) generated by product flows. The proposed profit sharing rule is then constructed from the dual of the profit maximization problem. It is both efficient and rational, with more fairness than the Owen set policy of classical linear production games.

The goal of this chapter is to provide a study of the coalition formation problem in supply chains using Hedonic cooperative games. The goal is to focus on the problems of (i) coalition structure generation, i.e., formation of coalition structures, such that agents inside a coalition coordinate their activities, but agents of different coalitions will work independently; and (ii) worth sharing, i.e., distribution of the worth generated by the coalition to its agents. We namely demonstrate that when cost-based proportional rule and equal allocation rule are used to divide the total created value, the efficient coalitions always exist and satisfy a set of desirable properties. Further; with the general results, we go deeper into a non-superadditive joint replenishment game with full truckload shipments for which we provide a polynomial algorithmic solution to generate the coalitions

In this chapter, we model the multiple unit single item procurement network formation problem as a surplus maximizing network flow cooperative game. Here, each edge is owned by a rational utility maximizing agent. Also, each agent has a capacity constraint on the number of units that he can process. That is, each edge can be assumed to have a capacity constraint on the flow that it can admit. The buyer has a demand for a certain number of units. The agents in the network must coordinate themselves to meet this demand. The buyer also has a specified valuation per unit of the item. The surplus which is the difference between the value generated and the minimum cost flow in the network, is to be divided among the agents that help provide the flow. We first investigate the conditions under which the core of this game is non-empty. We then construct an extensive-form game to implement the core whenever it is non-empty.