A lot can be derived from the voltage of electrolysis or fuel cells, if measured correctly and continuously. That has been the expertise of Smart Testsolutions for 30 years.
Interview: Eva Augsten
H2international: Mr. Neu, 30 years is a long time when things go up and down constantly, as they do with hydrogen. How did Smart Testsolutions manage to grow so old in such a volatile industry?
Wolfgang Neu: You have to have had a bad experience and learned from it. For us, it was the development of inverters for photovoltaics. Back then, as a young company, we developed an inverter to the point of functional maturity that had a better efficiency than all the others, and at that time we even installed the devices ourselves on rooftops. That was shortly before the 1,000 Roofs Program ended in 1992. After that, there was no market anymore. So we put our development on ice and focused on electronics for automotive applications, our core business. When the 100,000 Roofs Program was launched in 1999, we didn’t have a series-ready product, and we watched competitors like SMA grow rapidly. We didn’t want to repeat this mistake with hydrogen.
And what did that mean when applied to the hydrogen industry?
In 1996, we developed our first cell voltage monitoring system for the Daimler research department in Ulm. The first application was then in the NEBUS by Daimler-Benz, together with Ballard and later also Ford. Ten stacks were installed in it, each with its own electronics, interconnected using the master-slave principle in order to reach the required total voltage. That was very successful, led to an initial small series, and even to good revenues. There was a real hydrogen hype back then. But then Daimler, because of the political situation in the USA, saw no chance that the vehicles would be approved there, and the European market alone was too small. So a lean period followed. But after our experience with PV, we assumed that the topic would come back, so we had to hold out.
In photovoltaics, we didn’t have a series-ready product when the 100,000 Roofs Program was launched. We didn’t want to repeat this mistake with hydrogen.”
That means you lived off the automotive business in the meantime?
Correct, off measurement and test systems and engineering contracts. Cell voltage monitoring was initially nothing more than a voltage measurement assembly.
In parallel, there were research institutions and, from time to time, pilot projects, from VW, Ford, GM. We also developed systems for submarines, which were the first series applications, on average six to eight submarines per year. These were very special systems that we have continuously developed further.
In the automotive industry, Daimler then approached us again. At the time, however, it was still thought that cell voltage monitoring was only needed in the development phase, until the processes ran stably. Actually, today it is also firmly planned into series production. However, we no longer measure every cell, but for example the first and the last and then a few selected ones in the middle.
Initially, it was thought that cell voltage monitoring was only needed in the development phase. Actually, today it is also planned into series production.”
What is so complicated about the cell voltage anyway? Every multimeter from the hardware store measures voltage.
We have to deliver reliable data over years, at temperatures between −40 and +105 degrees Celsius, and also when the system is freezing or thawing. It can happen that so much moisture forms that the electronics are practically bathing in the fuel cell system. There are temperature shocks, there are vibrations, all of which we have to take into account.
What about space? In the automotive industry, everything is integrated into minimal space.
The very first stack from 1996 had 127 cells and was almost 50 cm long. Today you have 400 cells at half the length. Nevertheless, the measurement is faster and has more functions. In the beginning, the electronics still consisted of individual parts, then it became more and more integrated. Contacting is also an issue, you have to somehow get to the 1 mm thin cells in such a way that it holds even under vibration. In the beginning, they were still attached individually with tweezers…
But compactness has its limits. We’re talking about high-voltage systems with 400 or even up to 800 volts. That means we always have to maintain clearances and very high safety requirements so that the galvanic isolation of the supply and bus connection is absolutely safe.
With the electronics, we are now in the sixth generation. The cycles are becoming ever shorter.
Can one now speak of CVM as a series product?
The systems are intended for series applications, however, the applications are mostly in a prototype phase. The real series applications still fail because hydrogen and infrastructure are missing.
Speaking of which: What about electrolyzers? Do they also need cell monitoring, and is it easier? There is plenty of space.
We thought so too. But in contrast to earlier, today the electronics are usually wanted directly on the stack. So you’re operating in an environment with explosion protection requirements. And there are different bus technologies, more likely Profinet or Profibus instead of CAN bus like in automotive.
How is AI changing measurement technology?
For us, it doesn’t play a big role yet, but rather at the higher level, in data analysis. But there are approaches to integrate more intelligence into the cell voltage monitoring, and to also sample other sensors. We have already implemented something like this for a submarine. Now we are working on it for a customer in a significantly more integrated version.
And what about the competition from China?
So far, our Chinese competitors have hardly operated beyond their national borders. They are extending their feelers toward Japan, but the Japanese are very cautious, also because of political tensions. We also keep hearing that the Chinese products cannot keep up with ours in terms of stability and reliability. But in two or three years I expect competitive products from there.
So what are your plans to stay in the game?zu bleiben?
We are keeping our core business in measurement and test technology.
In hydrogen, there was a real slump starting in 2024. Now, in 2026, we are feeling a tentative revival, more projects are coming again for which we are developing series products. One is a funded project at a well-known supplier. In addition, we are in a B-sample phase with a major OEM, and in the offer phase for a customer from Japan. We are working on applications in aerospace, for example a prototype for a space organization, and for an aircraft manufacturer we are currently preparing fuel cells for flight operations. And with submarines, we now have three different manufacturers as customers. There are applications in practically every industry in which hydrogen is to be used.
Smart Testsolutions
About Wolfgang Neu and Smart Testsolutions
Wolfgang Neu is founder and managing director of SMART Testsolutions GmbH, headquartered in Stuttgart. His professional roots lie in automotive electronics development at Bosch, where he worked in advanced development on functional prototypes for the first direct injection. In 1988, together with two other colleagues, he founded an engineering office from which SMART Testsolutions emerged. Since then, he has shaped the company technologically and strategically with a clear focus on measurement and test technology for demanding electronics and automotive applications.
