Analysing approaches to specify automated manufacturing systems

Abstract

Automating manufacturing systems can achieve competitive advantage leading to growth in profits through efficiency gains and other advantages including safety of workers and quality of produced products. However, without accurate specification there is no guarantee of realising return on investment. Automated systems are becoming more complex as the need for customisability and variability of products increases and can only be satisfied through flexibility of production processes. To aid companies in specifying automation and mitigate the risks of project failure an approach is needed that guides users choices. The aim of the research was to investigate approaches to specify automated manufacturing systems to provide a basis for a methodology that would aid practitioners in this difficult task. The objectives were in two phases. Firstly to categorise and criticise conclusions of other researchers resulting in identification of themes and criteria for an approach. Secondly to experiment empirically with promising approaches in a company producing of automated manufacturing systems (AMS) and compare the results of the experiments with those found in literature and provide a ranking of themes and criteria to aid future researchers in designing new approaches to specify AMS. The methodology used was literature review followed by mini case studies in a host company to test theory. The results from literature and the experiments were classified into four themes quantitative modelling and simulation (QM&S), database decision aids (DDA), flowcharts and consultancy. These were compared using analytical hierarchy process (AHP) against the identified criteria; rapid application, usability by managers, considering costs and benefits other than financial ones, reducing required resources, being applicable to engineer to order products and usable at the early stage of planning. The results were the strengths and weaknesses of each theme defined by the identified criteria and showed that none of the themes fulfilled all of the criteria for an approach to specify AMS. For this reason a hybrid approach was proposed beginning with a flowchart group session to make an outline plan, followed by a database decision aid to provide options and guidance in creating a detailed plan. Finally, an optional simulation stage could test the planned system for suitability. It is hoped that the comparison of approaches will aid future researchers in the creation of new approaches to assist engineers in specifying automated manufacturing systems in a rapidly changing world.