ETH students accelerate private sector innovation

Innovation is critical for companies to remain successful over the long term. At the same time, innovation demands large investments and entails a lot of risk. "An innovation project can quickly cost CHF 2 million, and it takes a good two years until we know if the new idea can be translated into a marketable product with paying customers,” says Ian Roberts, Chief Technology Officer at the Bühler group. Bühler is a Swiss tech firm and globally leading producer of machinery for the food industry and other sectors.

Minimising the risk of investment decisions

How can companies accelerate their development and innovation processes while also minimising the risks involved? This question is at the heart of the research conducted by Mirko Meboldt, Professor of Product Development and Construction at ETH Zurich. Meboldt runs the university's Feasibility Lab, which supports companies in testing their ideas for innovation.

In this real-world laboratory, students and postdocs work together with clients to develop rudimentary prototypes that mimic the functions of industrial processes in need of change or improvement. When building these critical function prototypes, as they are called, everything that isn't essential for testing a certain hypothesis or innovative idea is omitted. The Feasibility Lab refers to this strategy as “lean de-risking”.

Last year, the Feasibility Lab and the Bühler group launched a collaboration called “BEXL – Bühler Exploration Lab” in order to test how the lean de-risking strategy could work in the company’s internal innovation process. Currently a team of eight students, led by three experienced project leaders from the ETH lab, is working at CUBIC, Bühler’s innovation campus in Uzwil. At the end of 2023, they presented their initial round of results to Bühler’s management.

Translating crazy ideas into prototypes

Can products that are processed with an extruder, such as animal feed or cereal, be dried with steam instead of hot air? From a physics perspective this method has the potential to be more energy efficient. To test whether this theoretical concept could be converted into a useful product, ETH students built a modular test facility in which different parts of the system could be quickly tried out.

They generated 160-degree steam with the help of an immersion heater that was connected to an oven. Here it is critical for the conveyor belts to deliver the material into the oven without any steam escaping. If any steam escapes, the gains in energy efficiency would be lost. The students needed two weeks to demonstrate that the method indeed works. They were also able to do so cheaply: they got the oven on sale for just CHF 70.

Practical application is key

“This is a great example of how our students can translate crazy ideas into prototypes that focus on obtaining key insights,” says Kai von Petersdorff-Campen, one of the three project leaders at the Feasibility Lab. "Measuring the results of these experiments creates a foundation on which the prototype can be developed into a real product in a series of sprints.”

Professor Meboldt is pleased that the methods from the lab also work in a real-world setting. “For us it is absolutely critical to work on real projects that show that new methods can achieve results,” he says, adding that private sector researchers also need to devote extensive attention to these methods to ensure their success. This works best if they have the opportunity to contribute to real projects, he adds.

Driven by curiosity

Students aren’t assigned any issues to work on from project leaders or professors. Instead, they are approached with ideas for innovation projects by the 24 business units at Bühler. Sixty projects have been set up to date, some rather simple and some more complex. Students have a lot of freedom to prioritise the projects as they see fit, with curiosity and expertise being the decisive factors.

The students work closely with engineers from Bühler when developing their solutions. First, they pester the engineers with questions until everyone has the same level of understanding of the issues behind the original idea. Depending on the complexity, the ideas are explored in different formats ranging from one-hour workshops to weeklong projects.

Recognising critical factors

One of the first business units that approached the students was responsible for grains and pulses. The issue at hand was developing a device that milling companies could use to determine the quality of oats when on delivery in order to determine an appropriate price.

The students’ first question was what exactly a device of this kind would need to measure. In addition to the grain to chaff ratio, the quality of oats is also determined by how easy it is to separate the husks from the grains, as this affects how easy they are to process. This means that the goal of the project was to translate these processing criteria into a small testing device.

“Here we were able to identify the acceleration of the grain as a critical function,” explained Arne von Hopffgarten and Diego Verzaroli when presenting their first prototype, which they crafted out of cardboard and powered with air pressure. Following several successful tests, they built a second prototype – using products from a hardware store – which was designed to mimic the de-husking process. Then they built a third, more sophisticated prototype, with de-husking results comparable to that of an industrial machine.

Lean de-risking: a proven approach

“This step-by-step approach is a great example of what we mean by lean de-risking,” says Petershoff-Campen. The effort put into an experiment should be healthily proportional to the current state of knowledge. Researchers refer to this as “the smart zone”.

For the first prototype, students needed just 18 hours, and they built it out of cardboard. The second prototype was ready after 40 hours of work. It was only after these successes that they started on the third prototype, which required an investment of 400 hours. With each step, the student team reduced uncertainty, which then justified investing more time and effort into the next prototype.

“I'm thrilled with how the ETH students have become part of the Bühler community, and I am very impressed with how quickly they were able to develop prototypes that provided answers to fundamental questions,” says Ian Roberts. “With our usual development methods, a project like the oat de-husker would have definitely taken two years instead of two months, and it would've burnt through a lot more money in the process.” The business unit is now developing a product based on the test results from the prototype, and they're expected to present it to their first customer this coming spring.

Opening up collaboration opportunities for other companies

The CTO of Bühler believes that smaller firms in the Swiss machinery industry could also benefit from a collaboration of this kind. “We’re currently reviewing the idea of opening up CUBIC to Swissmem partner companies to give them the opportunity to work together with ETH here in Uzwil,” he says.

ETH is also open to expanding the initiative. “We’re extremely pleased that our approach has also proven effective in practice and that it can be used to support industrial companies in Switzerland,” says Meboldt. The students are also excited, he says, and have gained a lot of insights through their real-world work.