KI and today? The lane departure assistant for industrial manufacturing

The topic of artificial intelligence is becoming increasingly important. The potential applications of AI seem to be almost limitless. The German government estimates that AI will account for more than one-third of future value creation by 2025.

What AI can already do today in industrial manufacturing was explained by our Managing Director Christian Friedrich on September 22nd at the forum “What can AI do?” of the promotional association of the Ruhr West University of Applied Sciences (HRW).

Until now, it has been a matter of extracting new information from data in conjunction with rules and heuristics, so-called business intelligence. Today, artificial intelligence and machine learning are used to generate new rules and logics from data.

The great advantage of this new methodology is that found rules and logics provide results even for previously not experienced and known states.

statmath as part of the ifm group

Since the beginning of 2019, statmath GmbH is officially part of the ifm group, which supports HRW, as a member of the non-profit support association. The ifm group provides its customers with the possibility to collect data with the help of sensors in the first place. However, the mere collection of data does not represent added value for companies. Only with the help of extensive data analytics and the scientific creation of algorithms, i.e. the development of artificial intelligence, forecasts can be made accurately for the future on the basis of historical digital traces and past data. In this way, it is possible to increase the profitability and performance of a company.

“As statmath, together with ifm, we are able to identify the needs in the market, cleverly capture sensor data and turn it into scalable solutions for the industry through AI,” explains Christian Friedrich.

The DataScience Toolbox

A big step taken by statmath and ifm this year is the integration of the first AI module in moneo, the IIoT platform for industry and manufacturing.

The DataScience Toolbox enables automatic monitoring, optimization of machines and equipment, and real-time assessment of manufacturing quality. Based on ifm sensor data and process parameters of the system, the SmartLimitWatcher, the first tool of the DataScience Toolbox, automatically detects anomalies at an early stage. For this purpose, a critical process variable (target variable) is permanently monitored with regard to the production quality or the plant condition (e.g. temperature, flow, vibration, current, etc.). The SmartLimitWatcher then uses the captured data to calculate a suitable mathematical model to describe the captured reality using common AI methods. The customer then still has the option of defining individual warning and alarm limits (low, medium, high), which are displayed in the application. In this way, it is possible to react quickly to deviations within the production process and to take appropriate proactive action through optimal early detection.

Monitoring of compressed air systems

Using compressed air systems as an example, Christian Friedrich explained the advantages of dynamic process monitoring using the SmartLimitWatcher. Compressed air systems are present in almost every manufacturing operation and are highly susceptible to the smallest of unnoticed leaks, plus they are among the most expensive forms of energy. As a result, AI-assisted monitoring of compressed air systems is crucial to efficiently maintaining overall manufacturing operations.