The CC-KING competence centre has set itself the goal of combining cutting-edge AI research with established engineering disciplines. This enables AI methods and Machine Learning to be utilized in accordance with the typical requirements and procedures of engineers. In classical engineering, for example, it is important to guarantee the behavior of complex systems. This issue was transferred to the application of AI methods in the project. Methods and procedures were developed that make it possible to obtain a statement as quickly as possible about the potential of AI methods in specific applications and the effort required to use them.
In order to clarify these questions as early as possible, numerous »QuickChecks« were carried out in the CC-King project. »QuickChecks« are a customized consulting service for end users offered by Fraunhofer IOSB, with which the use of AI and ML methods is examined and validated on a project-specific basis. The following is a small selection of QuickChecks carried out with the involvement of the Department Systems for Measurement, Control and Diagnosis. A complete overview is given here.
Our QuickChecks at a glance:
Automated anomaly detection in steam and gas turbine generator sets
In the course of the energy transition, large turbines will become less important and energy generation will shift to smaller, decentralised industrial turbines. This will necessitate a reduction in manual data evaluation. The extent to which an AI/ML solution can be developed to monitor steam and gas turbine sets was investigated.
Condition-oriented analysis of pipelines
When analysing the condition of pipelines, damage statistics are normally evaluated using conventional methods: damage characteristics are selected manually by experts and a statement is made about the condition of the pipeline system based on these characteristics. QuickCheck will investigate the extent to which an AI can assist the experts in selecting these characteristics.
Plant monitoring for grinding machines
In the future, the analysis of process data, for example for process monitoring and optimisation, will become increasingly important for both manufacturers and operators of grinding machines. This QuickCheck was used to analyse what information can currently be obtained from the process data of a system that reflects the current state of the art.
Early detection of anomalies in wastewater pumping stations
The aim of this QuickCheck was to recognise imminent pump blockages in wastewater pumping stations as early as possible. The main focus was on analysing operating conditions with high noise emissions and vibrations in surrounding buildings.
AI-assisted Train Control in Tunnel
The trams in the Karlsruhe light rail tunnel are controlled by a central train control computer, which is responsible for setting the points in the triangular track area, among other things. Due to the greater complexity of the train control system and the high volume of traffic in the tunnel, unexpected, short-term delays can occur. The QuickCheck analysed the extent to which an AI-based assistance system helps to support dispatchers.