A Model-based Approach to Improve Planning Practice in Collaborative Aerospace Design

The research reported in this paper complements previous work in the modeling and analysis of product development processes to reduce process risk and lead time. Such methods have found a receptive audience in industry, as evidenced by many academic case studies and the initiation of similar process improvement initiatives by the companies themselves. Given the ongoing ‘pull’ from a number of companies, we are interested to understand why such model-based methods have not yet impacted upon wider operational planning in industry. In this paper, based on an extended research study at a large UK aerospace company, we describe the application of the Signposting process modeling approach to the problem of developing a pragmatic method to support planning practice. The research described in the paper focuses on the elicitation, representation and manipulation of process information in a form which can inform planning and scheduling; during the study, addressing these issues was key to gaining support of the approach within the company. INTRODUCTION The research described in this paper is based on a longterm case study carried out in a large UK aerospace company. The study was initiated after the company carried out a ‘lessons learnt’ exercise following completion of an early phase of a major new development project. The exercise had highlighted a recurring issue thought to be of particular importance; the design-make plans for major components would detail manufacturing processes in great depth, down to the level of transferring components between job shops, whereas the design phase was represented at far poorer resolution – design tasks were planned on the scale of months instead of days. The programme manager sponsoring the research hoped to unpack these ‘black box’ design plans – it was envisaged that this would allow a detailed checklist of design activities and deadlines to be constructed and maintained as the project unfolded. This would lead to more effective progress monitoring, allowing schedule slippages to be identified and remedial action to be taken before an avalanche of consequences was triggered. After conducting some preliminary observations and informal interviews with the personnel concerned it was concluded that available planning and modeling techniques suffered substantial shortcomings in describing the behaviour of the complex design processes found in the company. In particular, the acyclic networks upon which the standard resource planning and Gantt-based scheduling packages are based are inappropriate to represent the iterative behavior of typical design processes. In such cases, uncertainty in the characterization, ordering and duration of design tasks is driven by the inherently opportunistic and responsive nature of designing. In practice, these uncertainties prevent the development and maintenance of a detailed, ordered list of design tasks and their expected durations. A new type of plan was required, and, just as importantly, a new way of approaching the planning activity – a more structured method which would be effective under the conditions of iteration and uncertainty about tasks and their ordering in the design process. This paper describes the development and initial application of such a method. OBJECTIVES The primary objective of the study was to develop a method to support more detailed task planning in the complex design process. This capability would improve the company’s