generic.de software technologies AG

LEWA pursued the goal of putting together an all-round carefree package for its globally distributed customers. How this vision could be translated into real services had to be worked out in the agile development process. Thanks to our many years of experience in agile and iterative forms of work, we were able to support LEWA as an agile coach in word and deed. Sprint by sprint, the optimal solution was created: a centrally controllable IoT platform that brings interpretable pump parameters directly to the user via a user-friendly customer portal.

Smart Monitoring for Predictive Maintenance

Smart monitoring is at the heart of the digitalisation strategy for LEWA pumps. In an interdisciplinary team, we translated LEWA's engineering knowledge, which has been built up over many years, into mathematical models and algorithms. This makes it possible for the first time to automatically interpret the determined pump parameters and derive diagnoses from them. The result is the early detection of undesirable conditions in the pump and the surrounding system, including localisation and possible correction. At the same time, the technology allows predictive maintenance - i.e. it predicts when which pump needs to be serviced.

IoT platform with Microsoft Azure

If the results obtained via Smart Monitoring are not reacted to in good time, even the smartest pump is of no use. And since LEWA pumps are used in the most remote places in the world, it is extremely important to keep an eye on the data continuously. For this purpose, we have developed an IoT platform based on Microsoft Azure, which continuously collects the data of the smart pumps. Depending on the customer's requirements, this data can be transferred directly to the cloud. The pumps, which are now capable of communication, have another advantage: the real data collected in bulk can now be consolidated for the first time. This means that LEWA has an overview of the field population at all times and can react more quickly and in a more targeted manner.

Customer portal with Digital Twins

The "other end" of the digitalisation solution is the customer portal. As a single touchpoint for customer self-service, it allows customers to look after their pumps regardless of location and time. All pumps can be continuously monitored and controlled via dashboards in the form of digital twins. Pump documents and spare parts lists are just a few clicks away. The nearest LEWA subsidiary and its sales partners are always consulted in order to keep service work and supply chains as fast and efficient as possible.

The devices from Wagner coat surfaces. As simple as this may sound at first, the subject matter is complex - especially in professional use. Because not all paints are the same and not all paints are the same. Each coating material has its own characteristics and requires the correct adjustment of the Wagner spray systems with regard to nozzle, dilution and pressure - the so-called coating parameters.

In order to obtain these coating parameters, painters and plasterers have to laboriously decipher the "small print" on buckets, click through the manufacturers' websites or download masses of data sheets. Here, J. Wagner GmbH wanted an automated solution that at the same time creates added value for any spraying systems.

Machine learning for material recognition

The plan was to be able to clearly allocate the material - e.g. a paint bucket - by scanning the EAN code. However, the reality is different. This is because the material manufacturers work with number ranges that are not centrally available, unlike commercial goods from the supermarket. The solution: a combination of intelligent image and text recognition. The paint bucket is photographed with the app and a machine learning algorithm recognises the respective manufacturer as well as the exact material type. From this result, the necessary parameters are read out from a database and displayed.

The Web Crawler

Besides the lack of the necessary EAN codes, however, there is also no manufacturer-spanning database that provides information about the coating parameters. Since the number of manufacturers, materials and products is hard to keep track of and new items are constantly being added, an automated solution had to be found: the development of a web crawler. In combination with Microsoft Azure Cognitive Services, it is also possible to determine important coating parameters from product data sheets (PDFs) that can be accessed online - currently a unique feature in this market segment.

Flexible API

Another challenge was to make an existing web-based backend API app-enabled. The solution was to develop a flexible API that can transmit all the necessary parameters with just one call and thus works both from web to app and vice versa. This also created the basis for future use cases, such as the connection of a customer portal.