Applying manufacturing batch techniques to fraud detection with incomplete customer information

Fraud detection is an important issue in service industries. Its technical challenges include the existence of complex patterns, its multivariate nature, and the incompleteness of records. As similar problems are observed in batch manufacturing process analysis, existing batch manufacturing techniques can be adapted and mapped to the fraud detection problem in the service sector. In particular, the batch library method can be modified to incorporate new information into incomplete customer records for analysis, while controlling the overall type I error rate. A real case is used to demonstrate the effectiveness of this approach, and simulation results are presented to show that the proposed approach consistently outperforms existing approaches.     

Analysis of multi-product manufacturing systems with arbitrary processing times

Multi-product systems with finite buffers and sequence-dependent set-up times are quite common in modern manufacturing industry. In practice, the distribution of machine processing time could be arbitrary, while in existing literature it is often assumed to follow an exponential distribution. In this paper, we develop an analytical method to study the multi-product manufacturing systems with non-exponential processing times. An embedded Markov chain model is constructed and two approximation methods, Gamma estimation and linear approximation, are proposed. The model is validated with high accuracy by numerical experiments and practical data from an automotive assembly system.     

Reliability estimation based on operational data of manufacturing systems

Maintenance management has a direct influence on equipment reliability and safety. However, a large portion of traditional maintenance models and reliability analysis methods usually assumes that only perfect maintenance is performed on the system and the system will restore to as good as new regardless of the kind of preventive maintenance work‐order that is performed. This is not practical in reality and may result in an inaccurate parametric estimation. The research objective of this paper is to develop a maximum likelihood estimation method to obtain more accurately estimated parameters based on the operational data of manufacturing systems, taking into consideration the difference between perfect and imperfect maintenance work‐orders. Weibull distribution is specifically studied for this purpose. A practical case study based on industrial operational data from an automotive assembly line is performed to illustrate the implementation and efficiency of the proposed reliability estimation method. Copyright © 2008 John Wiley & Sons, Ltd.     

Scheduling manufacturing systems with work-in-process inventory control: single-part-type systems

A real-time algorithm is developed for scheduling single-part-type production lines with work-in-process inventory buffers. We consider three classes of activities: operations, failures and repairs, and starvation and blockage. The scheduling objectives are to keep the actual production close to the demand, the work-in-process (WIP) inventory level low, and the cycle time short. A three-level hierardhical controller is constructed to regulate the production. At the top level, we determine the desirable buffer sizes and the target production level for each operation. At the middle level is a production flow rate controller that recalculates the production rates whenever a machine fails or is starved or blocked. The loading times for individual parts are determined at the bottom level of the hierarchy. The production scheduling algorithm is evaluated by using computer simulations for a variety of cases. Compared with a transfer line policy, a significant improvement in system performance is observed.     

Scheduling manufacturing systems with work-in-process inventory control: Reentrant systems

In this paper, we propose a procedure for production flow control in reentrant manufacturing systems. The system under study consists ofN machines and producesM product types simultaneously. Each part goes through the system following a predefined process and may visit a machine many times. All machines are subject to random failures and need random repair times. The scheduling objectives are to keep the production close to demand and to keep the WIP inventory level and cycle times at low values. The model is motivated by semiconductor fabrication production. A three-level hierarchical controller is constructed to regulate the production. At the top level of this hierarchy, we perform capacity planning by selecting the desirable buffer sizes and the target production level for each operation. A production flow rate controller is at the middle level which recalculates the production rates whenever a machine fails or is starved or blocked. The loading times for individual parts are determined at the bottom level of the hierarchy. Comparison with alternative control is made through simulation and it shows that the control policy performs well.     

Complexity in manufacturing systems, Part 1: Analysis of static complexity

This paper studies static complexity in manufacturing systems. We enumerate factors influencing static complexity, and define a static complexity measure in terms of the processing requirements of parts to be produced and machine capabilities. The measure suggested for static complexity in manufacturing systems needs only the information available from production orders and process plans. The variation in static complexity is studied with respect to part similarity, system size, and product design changes. Finally, we present relationships between the static complexity measure and system performance.     

超声衍射CT及其在不完全投影条件下的实验研究

超声衍射层析成像术是一种利用声波散射数据所携带的信息来反演物体内部结构的技术，具有广泛的应用前景。本文对这一领域的研究现状进行综述，慨述一些典型的超声衍射层析成像重构方法，同时探讨在残缺投影数据下的重构问题，对于不完全投影数据下的图像重构进行了实验研究。最后对实验结果进行讨论。     

Agent-based knowledge management system with APQP: implementation of semiconductor manufacturing services industry

Future challenges posed to semiconductor manufacturing services firms demands a system framework which could not only introduce, development, and design new products into mass production with customer-acceptable quality level and profitable approaches, but also re-use best practice in the shop floor to continuously optimise the existing order fulfilment process with effectiveness and efficiency. Advanced product quality planning (APQP) tool integrated with agent-based knowledge management (KM) intelligence develops a systematic framework that responds to such challenges. We propose an APQP tool working with an agent-based knowledge management system (ABKMS) which builds on organisational knowledge resources and works with the APQP concept as well as KM agents. The proposed framework reflects how an APQP tool with KM agents seeks significant inroads into process knowledge repository to recommend a solution for decision making in delivering semiconductor manufacturing services in customer selection, parts and material selection, bill of material, equipments, and tools selection, with a modified APQP concept to work out the approach to be used to develop the streamlined and common process.     

Scheduling manufacturing systems with blocking: a Petri net approach

This paper introduces a Petri net-based approach for scheduling manufacturing systems with blocking. The modelling of the job routings and the resource and blocking constraints is carried out with the Petri net formalism due to their capability of representing dynamic, concurrent discrete-event dynamic systems. In addition Petri nets can detect deadlocks typically found in systems with blocking constraints. The scheduling task is performed with an algorithm that combines the classical A* search with an aggressive node-pruning strategy. Tests were conducted on a variety of manufacturing systems that included classical job shop, flexible job shop and flexible manufacturing scheduling problems. The optimisation criterion was makespan. The experiments show that the algorithm performed well in all types of problems both in terms of solution quality and computing times.     

Synchronization in manufacturing systems: quantification and relation to logistics performance

The term ‘synchronization’ in manufacturing refers to the provision of the right components to the subsequent production steps at the right moment in time. It is widely assumed that synchronization is beneficial to the logistics performance of manufacturing systems. However, it has been shown that synchronization phenomena can be detrimental to systems in which they emerge. To study if synchronization phenomena also occur in and affect manufacturing systems’ performance, a formal quantification and holistic understanding of the types of synchronization phenomena emerging in manufacturing are needed. This article aims to fill this research gap by developing synchronization measures for manufacturing systems, applying these measures to real-world production feedback data and utilising them to test the assumption about synchronization’s beneficial effect on logistics performance. We identify two distinct synchronization types occurring in manufacturing systems, logistics and physics synchronization, and show that they are negatively correlated. Further, we show that logistics synchronization and due date performance exhibit anti-correlation and thus question the assumption that synchronization leads to higher efficiency in manufacturing systems. This article aids production managers in designing and optimising production systems, and supports further empirical research in production planning and control and production system design.     

Concept of integrated data processing in computer controlled manufacturing systems (FMS)

An adequate and economic automation of production plants demands standardized and transferable solutions especially as far as computer controlled production is concerned. New developments of hardware and software components offer potential users extendable systems which, depending on the degree of extension, are capable of coping with both the technical and the organizational information flow within the production plant. In course of research developments concerned with the automation of production installation, a DNC-system was. first built using standardized process peripherals (CAMAC) and modular process control software written in a high-level language (PROCESS-FORTRAN) and later implemented in industry.

In the current set-up are two complex flexible manufacturing systems, one for profile milling pieces and one for rotational pieces. These DNC-systems are being equipped with additional functions necessary for data processing as well as for material-flow, handling functions and process monitoring. The presentation deals with the development and set-up of these two systems.     

Performance evaluation of flexible manufacturing systems with blocking

Analytical approximations for the performance of flexible manufacturing systems (FMS) with blocking of machines due to limited local buffers are presented. The approximations are based on a detailed analysis of FMS configurations used in industry. The method proposed uses informations generated by applying the classical closed queueing network (CQN) model to the FMS. The approximations developed are tested against simulation models for a wide variety of FMS configurations. The results presented show that the approximations are very good.     

Analysis of automated guided vehicle configurations in flexible manufacturing systems

Automated guided vehicle (AGV) systems complement the operation of flexible manufacturing systems (FMS) by providing integrated automated material handling that capitalizes on the system's flexibility. Previous research considering AGV systems for use in FMS installations has focused on complex control strategies to reduce the congestion problem often encountered in these systems. Recently, attention has been given to tandem system configurations that reduce congestion and simplify system control. The present study uses the simulation methodology to compare the performance of three AGV configurations under a variety of experimental conditions. The results indicate that system size, load/unload time, and machine failure rate factors have significant impacts on the operation of the systems considered. In general, with respect to due date performance, it is recommended to use the traditional configuration in small systems while using the tandem/loop configuration in larger systems. Furthermore, it is shown that the addition of the loop to the tandem configuration mitigates the sensitivity of the tandem configuration to the load/unload time factor as well as significantly improving its performance under high load/unload times. Thus, if tandem configuration is desired to reduce congestion and simplify system control, investments must be made to directly reduce the load/unload times or to construct a loop to avoid the load/unload time penalty.     

Missing data methods in PCA and PLS: Score calculations with incomplete observations

A very important problem in industrial applications of PCA and PLS models, such as process modelling or monitoring, is the estimation of scores when the observation vector has missing measurements. The alternative of suspending the application until all measurements are available is usually unacceptable. The problem treated in this work is that of estimating scores from an existing PCA or PLS model when new observation vectors are incomplete. Building the model with incomplete observations is not treated here, although the analysis given in this paper provides considerable insight into this problem. Several methods for estimating scores from data with missing measurements are presented, and analysed: a method, termed single component projection, derived from the NIPALS algorithm for model building with missing data; a method of projection to the model plane; and data replacement by the conditional mean. Expressions are developed for the error in the scores calculated by each method. The error analysis is illustrated using simulated data sets designed to highlight problem situations. A larger industrial data set is also used to compare the approaches. In general, all the methods perform reasonable well with moderate amounts of missing data (up to 20% of the measurements). However, in extreme cases where critical combinations of measurements are missing, the conditional mean replacement method is generally superior to the other approaches.     

An approach to the dynamic modelling of manufacturing systems

Because of their complexity manufacturing systems are difficult to model. However, modelling is very often required in order to study the behaviour of the system. In this paper an approach is described, where an analogy is drawn between the behaviour of a manufacturing and a mechanical system. Manufacturing systems have to respond to a dynamic demand, namely, a demand that changes over time. Flexibility of a manufacturing system can be thought of as the ability and the rapidness with which the system responds to the dynamic demand. This resembles the behaviour of a mechanical system under the excitation of a force that changes over time. The paper attempts to establish a modelling method based on this analogy and uses this method in the study of a real industrial system.     

Using simulation and optimisation to characterise durations of emergency department service times with incomplete data

Simulation models of emergency departments (EDs) are often built based on incomplete data, for example, missing arrival times or service-time durations. The difficulty in collecting reliable and complete data can subsequently lead to invalid simulation results. To tackle this problem, we propose a simulation and optimisation method to characterise the unavailable durations of service times. Since many services in an ED are sequential and dependent on each other, this paper considers these multiple process steps cooperatively. We first use lognormal distributions to characterise the key service durations. Then we propose a new meta-heuristic approach, which combines an Improved Adaptive Genetic Algorithm (AGA) and Simulated Annealing (SA), IAGASA, to search for the optimal set of service-time distribution parameters. To address the difficulties of applying IAGASA when noise is involved in the performance measures and improve the simulation efficiency, we jointly apply IAGASA and Optimal Computing Budget Allocation (OCBA) technology. OCBA minimises the total simulation cost for achieving a desired level of probability of correctly selecting the best set of distribution parameters, which improves the search efficiency significantly. The experimental results indicate that our proposed method can find accurate estimates of service-time distribution parameters within a relatively short time.