Make-to-order bidding models with contingent orders and multiple customer classes : a revenue management approach | |
| Author | Bunthit Watanapa |
| Call Number | AIT Diss. no.ISE-04-02 |
| Subject(s) | Order Revenue management |
| Note | A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Engineering, School of Advanced Technologies |
| Publisher | Asian Institute of Technology |
| Series Statement | Dissertation ; no. ISE-04-02 |
| Abstract | This research aims to improve the make-to-order (MTO) bidding model that simultaneously optimizes the bidding price and due date under contingent capacity for each incoming order. The model is firstly generalized to incorporate multiple customer segments classified based on parameters of willingness to pay, sensitivity to short delivery time, quality level requirement, and intensity of competition. The winning probability function is also modified to be of a more practical and robust model in reflecting stochastic nature of customer's decision, while a simplified pattern search algorithm is proposed to improve the efficiency in searching for the optimal price and due date. Secondly, a further improvement on profitability based on the single-period optimization is explored through the impact of order resequencing. Sequencing rules, namely the EarlyDue- Date (EDD) for time-critical orders and First-Come-First-Serve (FCFS) are preliminary applied with the generalized model to determine the sequencing position of each incoming order. The research finds that the more flexible the resequencing can interact with capacity allocation and the expected tardiness cost of the arriving order, the higher the expected marginal revenue can be achieved. Since the sequencing problem to maximize the average marginal revenue earned per bid is NP-hard, a Genetic Algorithm (GA) is consequently proposed in searching for a near-optimal job sequence that compromises solution quality for less computational time. The model is also incorporated operational constraints and marketing policies to effectively reflect the interests of customers. Finally, we integrate a revenue management (RM) approach to the bidding model, aiming to improve profitability over a finite planning horizon. The approach leads to a dynamic price setting-dynamic inventory allocation class of RM problem. For each arriving bid request, the dynamic programming (DP) based RM model recursively computes, while considering forecasted future demand pattern, the expected accumulative revenues and particularly the estimated marginal value of production time slots required by the bid request. Using such insightful information, the firm can make decision on appropriate price setting, either to enter into the bidding with an optimal price-due date, or to defend the value of time slots by bidding at the time its estimated values or at the maximum price limit. Two decision rules, namely the dynamic price adjustment and overbooking are incorporated into the proposed RM bidding model to further enhance its yield. They are empirically evaluated their effect on and contribution to profits including efficiency of capacity utilization, to provide managerial implications in applying the model. To capture the dynamic and stochastic nature of real situations, simulation experiments are conducted to evaluate all proposed bidding models, when enabling multiple customer classes and order resequencing and integrating RM into the bidding model. Especially the RM model is evaluated on various scenarios that reflect different combinations of density of bid arrivals, aiTival pattern of time-sensitive orders, sequencing rule, and duration of considered time horizon. All experiments positively confirm that profitability of the MTO bidding system can be significantly enhanced as proposed by incorporating multiple customer classes, allowing flexible resequencing of the incoming job, and applying the RM approach. |
| Year | 2004 |
| Corresponding Series Added Entry | Asian Institute of Technology. Dissertation ; no. ISE-04-02 |
| Type | Dissertation |
| School | School of Advanced Technologies (SAT) |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Industrial Systems Engineering (ISE) |
| Chairperson(s) | Anulark Techanitisawad; |
| Examination Committee(s) | Bohez, Erik L. J.; Vatcharaporn Esichaikul;Kolisch, Rainer; |
| Scholarship Donor(s) | Royal Thai Government (RTG); |
| Degree | Thesis (Ph.D.) - Asian Institute of Technology, 2004 |