George Chryssolouris

Confidence interval prediction for neural network models

To derive an estimate of a neural networks accuracy as an empirical modeling tool, a method to quantify the confidence intervals of a neural network model of a physical system is desired. In general, a model of a physical system has error associated with its predictions due to the dependence of the physical systems output on uncontrollable or unobservable quantities. A confidence interval can be computed for a neural network model with the assumption of normally distributed error for the neural network. The proposed method accounts for the accuracy of the data with which the neural network model is trained.

Digital manufacturing: History, perspectives, and outlook

Abstract Digital manufacturing has been considered, over the last decade, as a highly promising set of technologies for reducing product development times and cost as well as for addressing the need for customization, increased product quality, and faster response to the market. This paper describes the evolution of information technology systems in manufacturing, outlining their characteristics and the challenges to be addressed in the future. Together with the digital manufacturing and factory concepts, the technologies considered in this paper include computer-aided design, engineering, process planning and manufacturing, product data and life-cycle management, simulation and virtual reality, automation, process control, shopfloor scheduling, decision support, decision making, manufacturing resource planning, enterprise resource planning, logistics, supply chain management, and e-commerce systems. These technologies are discussed in the context of the digital factory and manufacturing concepts.

A Comparison of Statistical and AI Approaches to the Selection of Process Parameters in Intelligent Machining

This paper presents an approach for the selection of a set of process parameters for use in machining control. The approach is aimed at providing a range of parameters within which machining operations can be optimized. Because of the complexity and somewhat unpredictable nature of the machining process, this approach combines process modelling with rule-based techniques. Modelling correlates process state variables such as surface roughness or chip merit mark to process parameters such as feed rate, cutting speed, and tool rake angle. The modelling techniques considered in this paper include multiple regression analysis, group method of data handling (GMDH), and neural network. A rule-based module determines the final operational range of control parameters based on user information and modelling predictions. The different modelling techniques have been evaluated using data from orthogonal cutting.

Flexibility and Its Measurement

Abstract Flexibility is a major factor for increasing the competitiveness of a manufacturing system. Flexibility, however, can not be adequately considered in the decision making process, if it is not properly defined in quantitative terms. In this paper different types of flexibility are presented and quantitative definitions of flexibility are discussed. In the context of this analysis, flexibility is defined as the sensitivity of the manufacturing system to change. The paper discusses also a quantitative definition of flexibility derived from the analogy of manufacturing and mechanical systems. The discussion includes industrial applications.

Laser Assisted Machining: An Overview

Since laser technology has considerable synergy with machining technologies, Laser Machining (LM) and Laser Assisted Machining (LAM) are relevant research topics. This paper attempts to give an overview of recent developments and research trends. Although scientific work on this area has contributed to the understanding of the process, there are still unresolved problems regarding the limitations of the techniques, optimum machining conditions, etc. The outcome of experimental investigations on LAM shows potential applications for this process but there are several issues to be resolved.

Dynamic scheduling of manufacturing job shops using genetic algorithms

Most job shop scheduling methods reported in the literature usually address the static scheduling problem. These methods do not consider multiple criteria, nor do they accommodate alternate resources to process a job operation. In this paper, a scheduling method based on genetic algorithms is developed and it addresses all the shortcomings mentioned above. The genetic algorithms approach is a schedule permutation approach that systematically permutes an initial pool of randomly generated schedules to return the best schedule found to date.A dynamic scheduling problem was designed to closely reflect a real job shop scheduling environment. Two performance measures, namely mean job tardiness and mean job cost, were used to demonstrate multiple criteria scheduling. To span a varied job shop environment, three factors were identified and varied between two levels each. The results of this extensive simulation study indicate that the genetic algorithms scheduling approach produces better scheduling performance in comparison to several common dispatching rules.

An Integrated Approach to Process Planning and Scheduling

Due to modern computer technology, large amounts of information can be available to the different levels of a manufacturing organization. The usefulness, however, of this information is very limited without adequate decision making methodology. Currently many decisions related to the operation of a manufacturing system appear in the form of process planning and scheduling. Because these two functions are performed separately, many problems arise due to conflicting objectives or unforeseen changes in the manufacturing system. Such problems can be minimized or eliminated with the aid of an integrated approach to process planning and scheduling. A common, integrating element of these two activities is the assigment of the various production tasks to the different factory resources (human beings or machines). The methodology described in this paper addresses the issue of assigning production resources as a multiple criteria decision making problem and suggests a decision making technique suitable for the manufacturing environment.

A virtual reality-based experimentation environment for the verification of human-related factors in assembly processes

Abstract This paper investigates the use of virtual reality (VR)-based methods for the verification of performance factors related to manual assembly processes. An immersive and interactive virtual environment has been created to provide functionality for realistic process experimentation. Ergonomic models and functions have been embedded into the VR environment to support verification and constrain experimentation to ergonomically acceptable conditions. A specific assembly test case is presented, for which a semi-empirical time model is developed employing statistical design experimentation in the virtual environment. The virtual experimentation results enable the quantification and prediction of the influence of a number of process parameters and their combination at the process cycle time.

An approach to operational aircraft maintenance planning

This paper describes a short-term planning methodology of the line maintenance activities of an airline operator, at the airports, during turn-around time (TAT). The proposed methodology supports decision making for deferring maintenance actions that affect the dispatching of the aircrafts, aiming at high fleet operability and low maintenance cost. Based on health assessment and additional information regarding operational and economical constraints at the operators fleet level, a multi-criteria mechanism evaluates a set of generated maintenance plan alternatives. An alternative is defined as the possible allocation of all deferred maintenance tasks to a set of suitable airport resources. The selected decision making criteria are cost, remaining useful life (RUL), operational risk and flight delay. A series of experiments is conducted in order to validate and test the approach.

The use of neural networks in determining operational policies for manufacturing systems

Abstract In a dynamic manufacturing environment, an operational policy prescribes when and how tasks are assigned to resources. Recent research has investigated the behavior of operational policies that can be tailored to specific manufacturing situations through the selection and evaluation of pertinent manufacturing decision making (MADEMA) criteria. Often, however, the relationship between the relative importance of these criteria and the overall performance of the manufacturing system is impossible to analytically establish. This paper explores the use of neural networks for identifying the relative importance of these criteria for given performance goals of the manufacturing system.