Knowledge retrieval in stamping process optimisation

Abstract

Knowledge, information, know how — these are some of the most valuable assets of a manufacturing organisation. Historically much of this has been stored in the minds and skills of individuals in those organisations. However, there are now many drivers to try and institutionalise such knowledge. Over the last 10 years many organisations have been through significant staff reduction exercises and in the process have lost valuable knowledge. Thus, knowledge capture, retrieval, reuse and adoption allows costly mistakes to be avoided and gives organisations a competitive edge in global operations. The significant question that needs to be addressed is how do organisations capture this information and knowledge from the design-manufacture cycle and, more importantly, how can it be retrieved for reuse in new product development? Previous work by the authors has looked at knowledge capture. This paper describes a new method that has been investigated to retrieve information-based textural tags of data instances, and how such retrieved information can be incorporated into optimisation of stamping of sheet metal parts to improve new product development and process design.

Keywords:

• Knowledge management
• Folksonomy
• manufacturing systems
• text categorization
• information retrieval

Additional information

Notes on contributors

J Sinclair

James Sinclair completed a combined degree in Systems Engineering and Information Technology at the Australian National University in 2002. Following that, he began work developing manufacturing information systems for the STAMP collaborative research centre. This work led to PhD studies in information architecture, which he completed in 2007. Currently, he works as interaction designer for ContentKeeper Technologies.

J Smith

Jeremy Smith completed a combined degree in Systems Engineering and Information Technology at the Australian National University (ANU) in 1997. He returned the following year to join the STAMP collaborative research centre at the Department of Engineering as a Research Engineer. He has spent the majority of the last 10 years working on industry-based research projects in the automotive and aerospace industry with Australian, UK and US firms. In 2008 he completed an MPhil at the ANU, continues to work on a number of external projects at the ANU and has also lectured on manufacturing technologies and systems engineering.

M Cardew-Hall

Prof Michael Cardew-Hall has had a career that has bridged academia and industry. In 1984 he joined the Rolls-Royce/Imperial College Teaching Company Scheme as an Associate, completing a PhD in Computer Aided Manufacture. He joined Rolls-Royce Aero Engines as a Manufacturing Systems Engineer in 1987 and was seconded as a visiting research fellow to the Dept of Computer Science at Curtin University in Perth, Western Australia, in 1989 working on CAD research related to turbine blade manufacture. On his return to the UK in 1991, he took on the role of Manager of the Process Modelling Department. He returned to Australia in 1993 as a lecturer in the newly-formed Department of Engineering at the Australian National University (ANU). Much of his research has been industrially orientated in the area of CAD, computer aided manufacture and modelling, and simulation and optimisation of manufacturing processes, in particular sheet metal forming. His current area of interest is the application of machine learning and knowledge-based systems to manufacturing, management of technology transfer and variability in assembly processes. At the ANU, Michael has held the positions of Head of the Department of Engineering, Deputy Dean(Education), Acting Dean of the College of Engineering and Computer Science, and is currently Director of Development and Innovation in the Office of Vice Chancellor.