Bits&Chips

Yieldstar significantly cheaper

Auteur: Mathilde van Hulzen en René Raaijmakers
20 december 2016 

Through its value engineering approach, Norma expects to achieve a significant cost reduction for ASML’s Yieldstar metrology system.

When it acquired engineering firm Mecon in 2015, Norma announced that value engineering was its business strategy. Production, manufacturability and cost now form the tier-one supplier’s core expertise. ‘We work very closely with our customers on engineering,’ says Norma’s CEO, Frank Biemans. ‘That enables them to better concentrate on their core business, while Norma focuses on manufacturability and cost. Everyone in the supply chain benefits from value engineering.’

In value engineering, a multidisciplinary team improves a product’s value by systematically analyzing all its features and deploying creative design techniques. By adding features, increasing performance or lowering costs, a product can be made to better meet customer needs.

Norma’s engineers are currently halfway through an intensive value engineering project for ASML’s Yieldstar. This metrology system checks wafers for defects in chip fabs. The technology is based on optical scatterometry. When the Yieldstar detects a defect, the device communicates that information to the wafer scanner, which adjusts its settings accordingly. That reduces the margin of error and raises the quality of the lithography right during production.

Norma has been contracted by ASML to manufacture and assemble the Yieldstar’s wafer stage module. The Veldhoven company isn’t the only player in the metrology market. It competes with companies such as KLA-Tencor and Applied Materials. In this price-sensitive segment, cost plays a significant role. That was ASML’s motivation for taking another good look at the Yieldstar’s design.

In the past year, Biemans says, a series of major improvements to the Yieldstar have been executed. ‘The results so far are amazingly good. We’re targeting a significant cost reduction of a few dozen per cent for the Yieldstar. What’s more, we hope to earn back the cost of the redesign in one year. We’re halfway through the project now, and it looks like we’re actually going to achieve our goals.’

Brainstorming sessions

Value engineering is the way to achieve a substantial reduction in cost, says Biemans. ‘For the Yieldstar, for example, it turned out that we could lower our specifications at the component level without reducing the whole machine’s performance. The Yieldstar has to meet stringent requirements, but sometimes those requirements have been too rigidly translated into parts specifications. That makes components unnecessarily expensive. By modifying the specifications and redesigning the components, you can achieve substantial savings. Examples include simpler or less assembly work, shorter machining times and the use of other materials.’

For the redesign of the Yieldstar’s mechanical parts, engineers from ASML, Norma and tier-two and -three suppliers sat down together at the drawing table. That collaboration was crucial, says Biemans. ‘ASML’s engineers know all about lithography and metrology; we contribute knowledge of production methods and value engineering. And our own suppliers understand specific components. Our success in reducing costs is truly the result of teamwork across the entire chain.’

Norma first organized a handful of brainstorming sessions with roughly six people. Those generated about two hundred ideas that were then analyzed for usability. ‘You have to dive in really deep in projects like these. If you decide to replace a material with a cheaper alternative, for example, you have to prove that the device’s static and dynamic properties won’t change. The customer decides which ideas are scrapped, because they don’t have a great enough return, for example, or they’re too risky. That happened a few times in this project.’

The new version of the Yieldstar now contains parts with a rougher surface, for example, so that less polishing is required. Other parts are milled instead of spark-machined because they don’t require that degree of precision. Some frames have been replaced by sheet metal versions. Specific balance masses are now made of cheaper materials with a greater volume, which led to the design of larger spaces inside the machine. And custom plugs and connectors turned out to be replaceable by standard parts.

A pneumatics supplier helped modify the system for pressure and flow measurements. ‘The air flow was previously measured in a fairly complex way using several different sensors, plus the associated electronics and software. One of our suppliers was willing to help us think up a new solution that only needs one type of sensor. That makes a pretty big difference in the cost.’

Smart manufacturing

Value engineering can be applied in many markets, says Biemans. It delivers the greatest value when it’s conceived as a process during product development.

‘Another important point is that value engineering helps developers learn to look at a design differently. They gain greater insight into the relationship between design, production and cost. Why would you spend a long time on complex component milling if you could also simply cast it and mill the result? You only have to realize that once, and understand it. The next time, you design the product better from the start. Value engineering is a method that can be learned.’

To fully profit from value engineering, design modifications must be executed quickly and efficiently. ‘So a project doesn’t grind to a halt, waiting for a signature from someone who’s on vacation,’ Biemans says. ‘If you want to make maximum use of value engineering, you have to make your business processes as simple as possible. For example, ASML has sped up its engineering change process, the way in which design changes are recorded and approved.’

In value engineering, it’s particularly difficult to come up with new ideas for a machine that’s functioning just fine. ‘The device has to keep doing exactly what it’s been doing and deliver the same performance, so as an engineer you have to understand all the system’s characteristics. Plus you have to know what you can and can’t fiddle with. Then you translate that into all those thousands of components, and then you look at which specifications can be relaxed and how you can manufacture parts better and cheaper. So you’re looking for smarter solutions that make the final product cheaper and preserve the profit margins of all the companies involved. That’s definitely a serious challenge.’

Norma and ASML now have a year of the two-year value engineering project for the Yieldstar behind them. Biemans expects additional significant improvements in the coming year. ‘We roll out accepted ideas in bundles, in a couple of iterations. That way, we don’t have to turn everything upside down and we can redesign our production process on time.’

Norma’s ambition is to excel at smart manufacturing and to build a close engineering relationship with its customers. ‘That’s worked out well for the Yieldstar,’ Biemans says. ‘It’s a textbook example of value engineering, and clearly demonstrates the enormous advantages that can be gained. That energizes me and motivates me to do the same with other customers.’

Abonneer direct op onze nieuwsbrief

abonneren