Ulrico Jansen - Lead Engineer

We always look for ways to improve what we are already doing well. Much of these improvements are data driven. In fact, we use data insights as part of our craftsmanship. Although we cannot share sensitive business details, we will gladly give you a glimpse behind the scenes.

As Lead Engineer, I’m responsible for managing our technical projects. You could say that I’m the link between our program managers and production. I deal with everything that goes on in between, from paperwork and first article inspections to supporting our CAD/CAM engineers. But I’m also focused on making improvements. This is where I get to combine my skills and expertise with my interest. Data is what makes the big difference when it comes to making real improvements.

What is the purpose of measurement data?

It is mandatory to measure and inspect all aerospace products to prove 100% product conformity. But where we make a difference is that we also have tools to analyze this data using statistics. We perform capability studies (Cpk and Ppk) which measure how well our processes perform in relation to customer requirements. We use Design of Experiments (DOE) to investigate and improve the corresponding process parameters. And we use correlation studies to select the key process dimensions and implement SPC charts on the shop floor.

Different dimensions

Here’s an example. Take a typical product, where it’s not unusual for each step of the process to have to take into account 20 to 50 dimensions: diameters, positions, radii, thicknesses, 3D form tolerances and so on. It’s a complex total consisting of all kinds of variables and correlations. Statistical analysis will help us decide whether we need to focus on the output variation or if we can simply compensate for the output by creating offset for this dimension. It can also tell us something about tool life, which is valuable information during the design and improvement of a manufacturing process.

How to get accurate measurement results

To be able to see and distinguish variations in a manufacturing process, you need to have accurate and precise measurements. Basic concerns like the correct maintenance and calibration of your measurement gages/equipment are important, but also how they perform in real life when operated by real people. That is why we conduct several MSA (Measurement System Analysis) and Gage R&R (Repeatability and Reproducibility) studies. It’s important to know which factors – such as operator, measuring equipment, procedure, environment, etc. – cause the widest variation and contribute the most to measurement uncertainty. Continually improving our measurement processes is necessary, so that we can keep up with the ever tighter tolerances in the aerospace industry.

How data makes the difference

Once you have all the data and know that it’s accurate, the next step is interpretation. And, with the many kinds of experts on our team, each expertise has its own needs when it comes to data presentation. You basically want the data to be translated to the expert; to be as concise and clean as possible. They should not have to spend a lot of time interpreting it or – even less practical – having to look up the figures they need. One example is the use of control charts with a predefined OCAP (Out of Control Action Plan). The operator will instantly see when his process deviates from the usual ‘common cause’ variation and when he should investigate the so-called ‘special causes’. This gets everyone to stay sharp and keep to an absolute minimum any variation in the output (dimensions). This not only creates products which fall within tolerance – as every customer expects – but also creates products which are as close as possible to the pre-defined target (nominal). This fits in perfectly with the lean six sigma philosophy and provides our customers with better products.

In the end, data contributes to our operators’ craftsmanship. We’ve always been keen on delivering the best products and we have our manufacturing processes to back that up. Aeronamic doesn’t just claim to be world class… we can prove it.