ACHEMA Start-up Award Finalist: PipePredict

When a hole suddenly opens up in the street, an entire city district is left without water or there is an explosion in a chemical plant, it can all have the same reason: Undetected weak points in pipes that lead to leaks or even pipe bursts. The startup PipePredict has declared war on such events.

If you search stock photography for the keyword "chemical industry," you inevitably come across images of endless pipelines snaking through plants and forming a seemingly inextricable tangle. BASF's Ludwigshafen site, by its own account, has 106 km of road, but 2850 km of pipes. While most of them run above ground in chemical plants and are therefore still relatively easy to monitor, many other pipelines are located in rough terrain or underground. Whether a pipe is corroded or otherwise damaged is usually only noticed when a leak has already occurred.

Christopher Dörner, co-founder of PipePredict GmbH, explains what such a thing means: "A very practical example of the problem we are tackling with our technology can be seen in the following picture. In this situation, a burst pipe washed out a street in Berlin Lichtenberg and the car parked above it sank into a hole of mud. And in the manufacturing industry, burst pipes or leaks also cause production stoppages and the escape of toxic and explosive liquids or gases." If leaks could be detected or even predicted earlier, much damage could be avoided. PipePredict GmbH wants to make exactly that possible.

Use what is there

The idea came about during a project at a water utility that co-founder Valerie Fehst was working on as a data scientist at an IoT startup. Particularly in the case of water suppliers, leaks are often only detected when noticeable water losses occur. Localization is very complex. Valerie Fehst used a new approach to data analysis that she had worked on during her studies. "With this new type of data evaluation, existing sensor technology in the network is integrated into the solution, thus minimizing investment costs. In addition, we use a digital twin to create a link between the real world and sensor data, and then evaluate the sensor data using algorithms we developed ourselves," explains Christopher Dörner. In this way, leaks can be detected at an early stage and pipelines can be monitored in real time without having to install additional sensors. This enables maintenance on the basis of predictive maintenance - i.e. condition-based, without the need for a failure to occur first.

Giant market opportunities

The three-person founding team of Christopher Dörner, Valerie Fehst and Tri-Duc Nghiem quickly recognized the potential of their solution: "When we realized that water losses of up to 50 % occur worldwide due to leaks and pipe bursts, we couldn't believe it at first. Deeper research showed that even here in Europe, losses of 25 % are not uncommon. In addition, we see immense potential in our technology and if we had not founded the company, this technology might never have been developed," says Christopher Dörner. The response from the market confirmed the impression that many companies were just waiting for such a technology. "We had initially assumed that it would be difficult to find the right contacts in large companies. In fact, however, this has turned out to be easier than we thought. Especially through founder competitions such as the ACHEMA Start-up Award, contacts are made with the right people from large companies. And if the representative present is not the right contact person, he usually knows the right person and makes the contact," says Christopher Dörner.