Abstract
Building locomotives in a foreign land with strict local requirements does have quality and delivery risks. Four (4) Orginal Equipment Manufacturers (OEM) were awarded a contract to manufacture 1064 locomotives from the state-owned enterprise, Transnet Freight Rail. This was subject to South Africa’s Treasury requirements of 55% local content. Developing a supply chain operation of this magnitude with local suppliers that may have little or no experience in manufacturing a specific OEM’s locomotive was a quality and execution risk. The local supply chain was to be bench-marked against established global sites that have decades of experience and knowledge of manufacturing locomotives.
Quality and on-time delivery for South Africa was not aligned to global targets. The delivery delays had a contractual financial penalty. Late deliveries and and over-run on operatings costs would have a negative impact on the project’s margin. To prevent incurring additional operating costs, one needed to perform the work to specification the first time, so there would be no need for rework costs.
Using Six Sigma’s Design Measure Analyse Improve Control (DMAIC) tool would statistically highlight the significant areas that were of concern. Once identified, these significant areas were to be leaned out. After a successful lean implementation, it would then be evaluated to prove that there had been a statistically significant improvement. This improvement could then be quantified into a percentage of cost avoidance. The inspection documents at the final quality gate provided a comprehensive database of the entire process by means of a ‘PASS’ or ‘FAIL’ critera. With only two potential outcomes, this data is considered to be attribute in nature; within a year into the program there were 94 data sets available to evaluate. Through statistical analysis it was determined that the data size was adequate, and the measuring system achieved the desired accuracy, repeatability and reproducibility. The defects were filtered through various tools for analysing, brainstorming and prioritisation to understand their inputs, i.e. root causes. The root causes that were determined to be significant were then resolved by applying lean principles. Implementation of lean to address these defects, combined with a new leadership approach to monitor and audit the new improved standard, resulted in an improvement of 15 percent. The process still showed a variation due to the omitted defects becoming significant after the initial problem areas were addressed. The use of Six Sigma and lean have a profound dependence on one another, and the application seems limitless. Six Sigma provides a focus with statistical clarity that lean cannot. However, with all the statistical clarity and without the use of lean principles, there would not be any improvements.