Warren W. Fisher

Modeling and Evaluating Service Quality Measurement Using Neural Networks

Effective measurement and analysis of service quality are an essential first step in its improvement. This paper discusses the development of neural network models for this purpose. A valid neural network model for service quality is initially developed. Customer data from a SERVQUAL survey at an auto‐dealership network in The Netherlands provide the basis for model development. Different definitions of service quality measurement are modelled using the neural network approach. The perception‐minus‐expectation model of service quality was found not to be as accurate as the perception‐only model in predicting service quality. While this is consistent with the literature, this study also shows that the more intuitively appealing but mathematically less convenient expectation‐minus‐perception model out‐performs all the other service quality measurement models. The study also provides an analytical basis for the importance of expectation in the measurement of service quality. However, the study demonstrates the need for further study before neural network models may be effectively used for sensitivity analyses involving specific dimensions of service quality.

An improved simulation model for cannibalization policy performance comparisons in a complex maintenance system

A model of a two-echelon (bases and a depot), two-indenture (machines have components which may have modules) maintenance system with spares, repair, cannibalization, preven tive maintenance, and two classes of manpower constraints is presented. A base is modeled with a SLAM II combined net work and discrete event simulation, while other bases and a depot are incorporated analytically. Although designed primarily for the study of cannibalization decisions, the model is also suitable for testing the performance of repair, spares, and other maintenance policies in complex systems, including those main taining fleets of aircraft. Applicability of the model and its new features are demonstrated by showing how previously-reported results change under different assumptions.

Analytic performance estimation for complex maintenance systems

Earlier models are extended and combined to provide an analytic means of estimating the average number of inoperative machines in a two-echelon (bases and a depot), two-indenture (machines have components which may have modules) maintenance system with spares and repair. Separate expressions are developed for the no cannibalization and complete, instantaneous cannibaliza tion cases. The expressions are suitable for verification of aircraft maintenance system simulation models as well as for estima tion of system performance and the maximum benefit of cannibalization.

Value stream mapping for improving academic advising

Value stream mapping (VSM) is a popular tool to identify opportunities for improving process efficiency. Originally designed to streamline manufacturing processes, VSM can also be applied to service processes. In this research, VSM is applied to academic advising of undergraduate students in a large department of a major university. Improvements to the advising processes are suggested following a VSM analysis of the current procedures.

Cryptography for computer security: Making the decision

Cryptography can be a powerful method of protecting data stored in or transmitted between computers, but it is not appropriate in all situations. This paper addresses the problem of deciding whether to use cryptography in a particular organization. After a brief description of the nature of cryptography, the circumstances making data protection desirable are described. The advantages and disadvantages of cryptography relative to conventional security measures are discussed, and a logical approach to making the decision of whether to use cryptography is presented.

Controlled release location of jobs in a hybrid of job and flow shops

The study investigates whether the location at which the order review and release (ORR) strategy is applied in a hybrid of job shop and flow shop influences its performance. Five different ORR strategies were applied at two different locations of the simulated shop in conjunction with four dispatching rules used at two shop utilisation levels and two due date tightness levels. The shop performance was evaluated by several job due date and shop related measures, including both mean and variability statistics. The results confirm that the effectiveness of an ORR strategy can be enhanced by controlling the location of its application. Furthermore, the ORR strategies based on the backward infinite loading (BIL) criterion outperformed those based on the modified infinite loading (MIL) criterion. With regard to the dispatching rule performance, the due date (EDD) rule yielded the best results on most of the variability measures, while the shortest processing time (SPT) rule did better on the mean measures.