Process optimisation

Citation

Antony, J. (2005), "Process optimisation", Assembly Automation, Vol. 25 No. 1. https://doi.org/10.1108/aa.2005.03325aaa.002

Process optimisation

Jiju AntonyDirector at the Centre for Research in Six Sigma and Process Improvement (CRISSPI), Caledonian Business School, Glasgow Caledonian University, Glasgow, Scotland, UK

Keywords: Process optimization, Response surface methodology, Taguchi

From an industrial standpoint, the purpose of doing any business is to achieve and sustain business profitability and to maximise return on investment (ROI). In an environment of increasing international competition where countries with lower production costs quickly catch up technologically, it is important to focus on maximising the utilisation of existing technology. In other words, rather than investing in sophisticated new technology, we need to optimise the use of existing technology as the former approach may be a very expensive solution. Process optimisation is a methodology for reducing production costs, reducing process variability, enhancing process capability and thereby improving product quality. The ultimate objective of process optimisation is to fine tune your existing process using existing technology and achieve more process yield with better product performance and quality.

Process optimisation is a systematic approach to experimental design which simultaneously optimises the time, effort and costs required in product and process studies for quality improvement. Complex systems or processes require investigation of key variables or parameters which affect the overall performance of the system or process prior to optimisation studies. This is called process characterisation.

The ability to optimize or improve a process is dependent upon the ability to control the process. The ability to control the process is dependent upon the access to reliable and valid measurements. A successful industrial optimisation thus entails a strategic approach encompassing the whole chain:

• •

  Measuring —> Controlling —> Optimizing

In process optimisation studies, one may be advised to answer the following questions (Antony, 1997):

1. 1.

   Which settings of the process or system under investigation give a product or process satisfying specifications?

2. 2.

   What settings of the process or system yield a minimum or maximum response (response is the output of a process or system)?

3. 3.

   How is a particular response affected by a set of process parameters over some specified region?

4. 4.

   What is the local geography of the response surface near the optimal (maximum or minimum) value?

In the strategy of industrial experimentation and process optimisation studies, one may need to consider the following phases:

^.Screening – the purpose of this phase is to separate out the “vital few” process or system variables from the “trivial many”. One may use Plackett-Burman designs, Taguchi Orthogonal Array designs or fractional factorial designs (FFD).

^.Characterisation and modelling – the purpose of this phase is to develop a regression model which explicitly illustrates the relationship between the response and a set of critical process or system parameters.

^.Optimisation – this phase uses a sequential approach to experimentation until an optimal location is found. Experimenters generally use response surface methodology (RSM) to determine the optimum operating conditions of the process or system. This method is more powerful in seeking the optimal condition as opposed to traditional “one-factor-at-a-time” (OFAT) or Taguchi Methods (TMs) of experimental design. These two methods will produce results that are only 60-80 per cent of the total improvement to be made as opposed to over 80 per cent using RSM (Schmidt and Launsby, 1992).

Although process optimisation studies have been around for many years, it is not widely accepted and applied by industrial engineers and quality professionals in the UK manufacturing organisations. Experimenters and scientists in the UK are far more likely to use their “home-grown” solutions for process optimisation problems. Interesting enough these “home- grown” solutions are in accordance with OFAT approach to experimentation as our managers are after quick-fix solutions which yield short-term benefits to their organisations. The above problem should be tackled at the grass-root level by educating our engineering and business graduates at the university level about the importance of experimental design (ED) and process optimisation studies for improving product quality and process effectiveness. I truly hope our young engineers and business graduates will carry out research in these areas that will bridge the gap in the knowledge, expertise and skills required by them in their later careers for problem solving activities in organisations.

Optimisation vs investment in technology: a personal experience

• •

  During my endeavours in learning the ropes of process optimisation, I came across an interesting scenario with a pen manufacturing organisation. This organisation was experiencing a very high customer returns problem (about 35 per cent) due to low mean strength and excessive variability in strength of plastic pen shells. After an informal discussion with the senior management team, I was surprised to find that they were thinking of investing in new raw material and technology as a solution to this problem. A few well-placed questions brought to light the fact that they had never attempted to optimise their core production processes. Though I considered myself “green” at that point of time, I bravely faced the senior management team with my option of process optimisation as opposed to their extremely expensive solution. An application of a well designed experiment along with RSM helped me to understand and optimise one of their core production processes. The mean strength of plastic pen shells improved by 30 per cent and variability in strength was reduced by 18 per cent at the expense of less than $3k. The organisation saved several hundreds of thousands of US dollars and I learnt a valuable lesson from this case study – “Optimise your process before opting for expensive new technology”