Whereas machines, systems and components were previously operated in isolation or directly in connection with control software, in the age of the Industrial Internet of Things (IIoT), systems are given access to the internet and thus a kind of additional digital identity. Users can interact with the devices in real time, retrieve data and use it based on a combination of analytical methods and predictive algorithms. Not only does this result in more efficient operating processes and lower costs thanks to the possibility of preventive maintenance, but procurement planning can also be optimized.

However, companies that want to achieve long-term added value on the basis of IIoT should rethink their economic activities and implement data-based business models. This is underpinned by a study by the Capgemini Research Institute, which found that companies that lead the way in data use generate up to 22% more profit than their competitors.

Springboard for digital change

Using predefined templates and function modules, users from different departments can quickly build their own data apps based on machine data and offer digital value-added services for their products, such as visualizations of important key figures, monitoring dashboards for 24/7 monitoring or complete machine learning workflows. Thanks to a visually appealing user interface, they can combine the individual elements into a wide variety of applications and integrate them into existing systems, similar to a modular system.

Integration into IT systems

Platforms with an architecture designed for low-code throughout are particularly suitable for the implementation of IIoT projects. As the platform is used by a large number of users, the software should be created in many small parts and then be able to be combined into a functioning whole. Similar to conventional tools for software creation, low-code platforms also contain a data memory, an integrated development environment (IDE) and a development language. They also offer several APIs - REST or GraphQL-based programming interfaces have proven their worth, allowing the platform to be seamlessly integrated into IT systems from a wide range of manufacturers.

One of the core functions of every low-code platform is visual development, which allows the processes and their data models to be displayed. The machine data is read directly from the database via graphical operations, which is automatically integrated into the platform. By converting the data models into SQL queries and relational tables, the data from external APIs can be made available to users automatically even before the application is started.

Supported by functions such as data filtering, grouping, sorting or aggregation, users can carry out an initial statistical analysis independently. They then switch to a Data App Builder to create their own data-based apps using parameterized dashboards and graphical widgets. The in-depth analysis of the machine data takes place in a special editor, in which formula elements can be created and edited in a similar way to Excel, but here on the basis of an intuitive graphical user interface.

AI-supported applications

For precise predictions, low-code tools can also be used for machine learning applications, provided that the IIoT platform used offers product-internal, real-time capable, AI-supported development. Users can use graphical interfaces to create and train their own machine learning algorithms and later identify typical patterns in the alarm data.

The author: Michael Breidenbrücker is the founder and CEO of Senseforce in Dornbirn, Austria.

© Senseforce

In addition to 24/7 real-time monitoring, companies currently use low-code platforms primarily for the predictive maintenance of their machines, plants and other systems. By comparing the machine and system data recorded during operation with other data, such as idealized models, the software detects faults and malfunctions as they occur - often long before the malfunction actually occurs. Trained users can draw conclusions about existing faults in the system from the recorded noises, speeds or temperatures and, in the event of a service call, work in a targeted manner to rectify the fault on the basis of the collected system data. The estimated time of failure of a component, such as a seal or bearing, can thus be precisely determined, which not only reduces maintenance costs but also the failure rate of a machine or component.
For example, service staff can use a simple graphical interface to compile monitoring boards for their applications using drag-and-drop. These bring together the measured values recorded by sensors in the networked systems - such as temperature, vibrations, material jams, capacity utilization or wear. For the evaluation, product and service experts can define limit values that must not be exceeded or fallen short of. If this is the case, the machine sends an automatic notification by email or text message.