The solution enables companies to connect machine data with a mouse click, harmonize and integrate data, or run visual analyses. Riemer and his comrades-in-arms founded without a product, they fully rely on open source and develope modules and services around Streampipes that customers have to pay for. At Hannover Messe, they will show their solution in Hall 12, Booth E39.

Open source approach

Machine builders and plant operators can work freely with the Streampipes solution and run initial analyses. At the same time, the community provides a great deal of input for the further development of the application and also for modules. The goal: The application is intended to help make IoT data analyses in the production and logistics environment accessible even to less technically savvy users. The goal is to use Streampipes to offer the simplest solution to analyze continuous data streams in the IoT environment and to offer this in the form of an open-source solution that can run completely without a connection to an external cloud provider. For this purpose, machines or other data sources can be connected within minutes. In a graphical editor, the system then provides an expandable construction kit of algorithms and functions that make it possible to evaluate the data independently. Riemer and his colleagues rely on an adapter for the semantics of the data, which supports 30 protocols to date, and then merge in data at a higher level.

Building a knowledge platform

Unlike other no-code solutions on the market, the Bytefabrik.AI founders focus on interaction. The Karlsruhe-based company wants to bring together the knowledge from the many minds in manufacturing. Datastories is the name of the module they offer their customers - for money. Many plant operators find patterns in the data, but it always needs the domain expert. "Companies don't have the one expert, though. There is the maintenance engineer, the machine operator, the quality manager or the process manager. In the past, they would sit down in meetings and look at the patterns and make decisions. We automate that with the Datastories," Riemer explains. If the customer sees something in the data, he can add questions to the results. The user then sends the Datastory to different domain experts in the company. They can then dig deeper into the data because it's all stored and attached in the system and give their feedback. "We pull the knowledge together, build a knowledge platform for the company, and also use the results to train the AI models." This is how an open source project becomes a business model for a startup from Karlsruhe.