Asia-Pacific Journal of Operational Research

In this paper we study an M/T-SPH/1 queue system, where T-SPH denotes the continuous time phase type distribution defined on a birth and death process with countably many states. The queue model can be described by a quasi-birth-and-death (QBD) process with countable phases. For the QBD process, we give the computation scheme of the joint stationary distribution. Furthermore, the obtained results enable us to give the stationary queue length distribution for the M/T-SPH/1 queue.

A two-machine flow shop scheduling problem with rejection is considered in this paper. The objective is to minimize the sum of makespan of accepted operations and total penalty of rejected operations. Each job has two operations that could be rejected, respectively. Operations on the first machine have penalties α₁ times of their processing times and operations on the second machine have penalties α₂ times of their processing times. A -approximation algorithm is presented for the case where min{α₁, α₂} < 1 and max{α₁, α₂} ≥ 1. A dynamic programming algorithm is provided for general α₁ and α₂. A fully polynomial-time approximation scheme (FPTAS) is designed for all NP-hard cases.

In this paper, we study a discrete-time finite buffer batch arrival queue with multiple geometric working vacations and vacation interruption: the server serves the customers at the lower rate rather than completely stopping during the vacation period and can come back to the normal working level once there are customers after a service completion during the vacation period, i.e., a vacation interruption happens. The service times during a service period, service times during a vacation period and vacation times are geometrically distributed. The queue is analyzed using the supplementary variable and the imbedded Markov-chain techniques. We obtain steady-state system length distributions at pre-arrival, arbitrary and outside observer's observation epochs. We also present probability generation function (p.g.f.) of actual waiting-time distribution in the system and some performance measures.

Due-window assignment and production scheduling problems are important issues in operations management. In this paper, the problems of common due-window assignment and scheduling of job-dependent deteriorating jobs and multiple deteriorating maintenance activities simultaneously on a single-machine are investigated. We assume that the machine may be subjected to several maintenance activities during the planning horizon. However, due to the restriction of budget of maintenance, the upper bound of the maintenance frequency on the machine is assumed to be known in advance. Moreover, we assume that the duration of each maintenance activity depends on the running time of the machine. Polynomial time algorithms for the all the studied problems are provided, where the objective is to simultaneously minimize the earliness, tardiness, due-window starting time, and due-window size costs.

Hospitals must maintain safe nurse-to-patient ratios in patient care units to offer adequate and safe patient care. Since the patient demand is highly variable, during high patient demand periods temporary or overtime nurses are hired to ensure safe nurse-to-patient ratios. These overtime nurses incur higher expense, and are often less effective. We study the problem of permanent nurse staffing level estimation under demand uncertainty as a newsvendor model. Our models are based on limited moment information of the demand distribution. Additionally, we introduce the use of asymmetric cost functions representing overstaffing and understaffing nursing costs. Findings using data from the general surgery and intensive care units at hospitals in Chicago, IL and Augusta, GA are presented. Computational results based on publically available cost data show that 3.1% and 7.3% annual cost savings result by introducing salvage value and newsvendor optimization in intensive care and general care units respectively. This new staffing scheme also improves patient safety as shifts are staffed with more permanent nurses.

For shipbuilding, spatial scheduling and workforce assignment are two important issues for the operation management in block assembly shops. Spatial scheduling is to decide where and when to assemble blocks, while workforce assignment is to assign working teams to the blocks. Traditionally, they are made separately at different levels in the production management framework. The combining of such decisions presents additional complexity and new problems. This paper proposes an approach that solves jointly the spatial scheduling problem and the workforce assignment problem. The objective is to improve the coordination among working teams and increase the productivity of assembly shops. A spatial layout strategy is designed according to extreme point and deep bottom left strategies. By using genetic algorithm, a solution method is developed based on the spatial layout strategy and several assignment and sequence rules. Some computational experiments are conducted to evaluate the performance of the presented algorithm and compare it with other commonly used methods. Computational results validate the effectiveness and efficiency of the proposed algorithm.

This paper explores the operational and market efficiencies of various biotechnological company formats in Taiwan during the period 2006 to 2008 by adopting the two-stage directional distance function and the meta-frontier concept. It conducts an efficiency assessment by taking into account a five-year time lag between R&D investments and actually obtaining a patent in order to better reflect the actual business operating process and thus more accurately gauge the efficiency of business operations in biotechnological companies. The empirical results show that non-listed companies registered better operational efficiency in the first stage, whereas their market efficiency in the second stage is inferior to that of listed companies when the meta-technology set is used as a basis. The average operation gap ratio and market gap ratio of non-listed companies exceeded those of listed firms, indicating that the technological heterogeneity (i.e., the gap) of non-listed companies is smaller than that of listed ones.

This paper uses set theory to analyze cellular manufacturing systems. We develop a structural architecture to investigate the previous literature. We divide all research problems into three types — what, why, and how. We advocate studying these three types of problems from two managerial perspectives — philosophy and science. To fully and deeply understand cellular manufacturing, both three-problems and two-perspectives are important. We use the developed architecture to review cellular manufacturing literature and point out the weakness in the previous studies. This review motivates the research of this paper. We use simple but rigorous set theory to analyze two what-problems. Underlying concepts, such as part family, machine cell, cellular manufacturing system, bottleneck machine are defined. Following these definitions, we deduce sufficient and necessary conditions for perfectly partitioned cellular manufacturing systems. Two sufficient and necessary conditions of perfect partition have been discovered — prior perfect partition theorems check the potential independent manufacturing cells within an arbitrary manufacturing system; in contrast, posterior perfect partition theorem verifies a perfectly partitioned cellular manufacturing system. To our knowledge, we are among the first to apply rigorous mathematical models to these two types of what-problems for cellular manufacturing. We try to provide mathematical vocabularies to discuss common problems inherent in the design of cellular manufacturing systems.

In a recent paper, Ozturkoglu and Bulfin (Ozturkoglu, Y. and RL Bulfin (2011). A unique integer mathematical model for scheduling deteriorating jobs with rate-modifying activities on a single machine. The International Journal of Advanced Manufacturing Technology, 57, 753–762.) formulate a unique integer program to solve the single-machine scheduling for the objectives of minimizing makespan and total completion time. They also propose efficient heuristic algorithms for solving large size problems. However their heuristics are not optimal and so the NP-hardness of the considered problem is still open. In this note, we show that a more general problem can be optimally solved in polynomial time. We also provide optimal polynomial-time solution algorithm for the parallel-machine case to minimize total completion time.

This note first succinctly summarizes the currently available methods to solve the various nonconvex free disposal hull (FDH) models for technical efficiency as well as for minimum costs. It also offers some empirical illustration as to their computational efficiency. Second, this note briefly points out that the recent article by (26) and its correction by (4) proposing an extended enumeration method to solve for technical efficiency evaluated relative to this family of FDH models contain no original results.