How to make a design right first time: “The solution is always process-based”

Rapid technological progress and rising demand for electronics pose major challenges to the electronics sector. Because how can you do this in a world of component and personnel shortages? time to market shorten the time, increase production and at the same time meet the increasingly strict technical requirements of the customer? Right first time, a method in which a product rolls out of the factory in one go without errors, no longer seems an option but a must. Dick van Hees (MRMc) and Ruud van Beek (ASML) share the plenary lecture at the D&E event their experiences with right first time and dare to name the pain points.   

We speak with Ruud van Beek, Technical Supplier Manager Electronics at ASML. Twelve years ago, he and his colleague Dick van Hees were at the forefront of the change in mentality at the Dutch tech giant that led to the right first time-method. A method that was still quite new and unknown in our country at the time. Van Beek: “Fifteen years ago we purchased components externally and then assembled them ourselves. The failure rate was extremely high, at its lowest point even around 60%. It was clear that things could no longer continue like this. That is why we have completely changed course; by plug 'n pray Unpleasant plug 'n play”.

Resistance

"Bee right first time it is about delivering products to the customer that are immediately of excellent quality, so with a very low error rate. This only works if you involve suppliers and designers in a project from day one,” Van Beek continues. In practice, this turned out to be easier said than done. “When we started working on this twelve years ago, we faced resistance. Colleagues were used to a certain working method and did not want to simply change it. We had to convince them by the benefits right first time results for our customers and also for ourselves. Job satisfaction really increases if your products keep getting better.”

Dissatisfied

Retired ASML employee and advisor Dick van Hees still remembers this exciting period like it was yesterday. “Ultimately, as a company you only have the right to exist if your customers are satisfied and they were not at that time. That is why we immediately took a major approach by setting up a project with three OEMs from the area and seventeen of our joint manufacturing companies.”

The project team enlisted the help of independent research agencies, including IMEC and TNO. The goal was to develop a model with which the error rate could be predicted based on existing data. The government made a subsidy available and a year and a half later the model was ready. Van Hees: “I am still proud of the result, because the model meant a huge improvement for our business processes. If you know what went wrong in the past, you can make predictions and take measures to prevent mistakes in the future.”

Easy principle

“The principle is actually very easy,” Van Hees explains, “if you put the parts in a BOM (bill of materials) provides insight, you can use the BOP (bill of process) generate from. With past data about the capability of a process, you can then predict how many products will fail when delivered. The next step is to reduce this number and various solutions are possible.”

“During my presentation I will give a number of practical examples,” Van Hees continues. “Suppose a technician forgets an O-Ring during production, which could cause a machine to leak during use. A possible solution is to only make the required O-rings available in the production order (sealant). The first step is always to see whether the design can be made in such a way that no errors can occur. Step two is to design the process in such a way that it is impossible to make the mistake. Step three, which you really don't want to do, is adding a test that catches the error before it reaches the customer. We outsource all these steps to our partners, which allows us to focus even better on our specialty: making machines.”

Around the table

Van Beek. “It is often the small details that determine whether a design is successful. The customer is not interested in the fact that the cause of the problem is a forgotten screw. All the customer wants is for the machine to work. And that is understandable, ASML's latest generation of chip machines cost more than 100 million euros each. That is why we continuously look at how we can further optimize and tighten our business processes. A chain is only as strong as its weakest link and that also applies to our machines.”

Closing the feedback

Finally plays closing the feedback an important role in improving operational processes within ASML. Van Beek: “Dare to face mistakes from the past and incorporate them into your design. This way you avoid making the same mistakes over and over again.” Van Hees agrees with his former colleague: “With everything we design, we ask the question: is it feasible, testable, unambiguous and complete. A certain margin of error is acceptable to the customer, for example 1 in 1000, but as a company you must substantiate this. Testing and design errors are often the result of minor human imperfections, but ultimately the solution is always process-based.”

Come to the lecture

Would you like to learn more about the experiences and best practices from ASML with right first time? Then come to the plenary lecture by Dick van Hees and Ruud van Beek at the D&E event. Admission to the lecture and event is free, but you must register in advance via the website of the event. See you in Den Bosch!