Backed up by 40 years of experience, the Selha Group is a manufacturer specializing in the manufacture and production of electronic and wired printed circuit boards, and PCB assemblies. IT has 900 employees spread across 4-production sites in France and Morocco working to support the Group's manufacturing activities, and aims to respond to the challenges of tomorrow. The Selha Group offers low- to medium-volume production of high-quality electronic components to respond to demanding technical requirements and offer superior quality. The Group’s high manufacturing standards and its wide scope of service are intended to cater to manufacturers in aeronautical, telecoms, energy, medical, and creative technology sectors.
In the Selha Group, we have been carrying out statistical research into the operating capabilities for the last 15 years. Our team of statisticians has already carried out around 900 successful analytical studies. The purpose of these programs is to establish whether the manufacturing processes are capable of delivering components that satisfy our clients' demands. By “capability”, we mean the relationship between delivery timescales (what the client needs) and actual dispersal (what we can manage). It compares the performance of a process (in other words whether it is capable of delivering) with tolerance limits in order to optimize product reliability. For example, did you know that we have discovered that, overall, one measurement in ten does not meet this capability standard in a test environment?
The challenge that the Selha Group faces is to find a solution to examine the capability of newly developed products emerging from the aeronautical sector and to more rapidly enhance their reliability before launching them onto the market. The larger the sample size, the higher the level of precision, and the lower the error margin of the capability estimates indicator. However, in the aeronautical sector, the quantity of prototype batches is typically around 10 products. So how can we help customers bring their products to market more quickly?
Thanks to its many years' experience, the Selha Group, has a solution, presented by Gilles Chassat, director of Quality, Hygiene, Safety, Health, and Environment:
Minitab: What is the solution you have developed?
Gilles Chassat: We have implemented a statistical solution to enhance the reliability of aeronautical products with batch sizes of 10 components. With small sample sizes, you are bound to see uncertainty and a higher margin of error. We have decided to include this uncertainty in our statistical reports. In the past, we used to work in a binary mode with just one criterion: either the product was capable or it wasn’t. Now we’ve moved to a system with three levels: the area of non-capability is shown in red; the area of uncertain capability in yellow and the area of proven capability in green. Below is a diagram showing an initial target of 1.33.
When the CPK (process capability index), is greater than or equal to 1 and less than 1.67, the process capability is uncertain; we are in the area of uncertainty.
When the CPK is greater than or equal to 1.67, we have a capable process.
Minitab What actions do you take in cases of non-capability (red) and this area of uncertain capability (yellow)?
Gilles Chassat: Relying on the completeness and accuracy of these analog measurements, we go back to our technical management who will then recommend either an up-front solution to our clients and partners or suggest that we work together to solve the problem. There is a range of different actions that can be applied to the product, interface, design, the test plan or the components. We discuss these with the client in order to get as many measurements as possible into the green capability area.
Minitab: How has Minitab contributed to the solution?
Gilles Chassat: We have been using Minitab for a number of years and it has become a real benchmark within the field. It’s a genuine guarantee of safety in the manufacturing sector we operate in; we can count on its reliability and performance. We acquired this product a few years ago and there are countless examples of why we put our trust in it. The starting point for the solution that we are offering came from Minitab. During the summer of 2018, as Director of Quality, Hygiene, Safety, Health, and Environment, I found myself confronted by the problem of how to reduce the time to market for our aeronautical products. While putting together a solution, I began to use the Minitab tool. I was inspired by the R&R (Repetition and Reproduction) method that Minitab used for statistical analysis, which makes use of a 3-tier system. Although the data and the way it is used are totally different, it was a source of inspiration, which allowed me to adapt and build our methodology.
Minitab: Can you give me a concrete idea of the results you've achieved?
Gilles Chassat: The primary objective is to enhance the statistical reliability of the product. This should result in mitigating the risks involved in repair work (additional tests, component change, etc.), Improving output and reducing testing volatility. For some clients it is a paradigm shift, but after initial trials, they find they can’t live without it. In one specific instance, a client managed to achieve a return on investment in the space of one month. Thanks to statistical research, they produced just two versions of the prototype instead of the three they had originally planned. This alone was enough to help them make savings of €100,000 in one year. Right now, we estimate that we are saving 30% per client of the costs associated with the prototype process.
Using Minitab, Gilles Chassat and his team are able to process Cp/Cpk capability statistics. Above you can see an example of the sort of analog measurement used to generate the results dashboard necessary for a batch size of 10 components.