AI-Based Error Management in Value Chains
New WZL Research Project “Value ChAIn” Investigates Fault Correlations in a Cross-Value Chain Context
In order to meet the demand for commercial vehicles, a production network of suppliers and original equipment manufacturers (“Original Equipment Manufacturer”, short “OEM”) was created in Germany, which together with transport companies covers the entire value chain. This production network has been able to establish itself as a key sector of the German economy. However, the production of commercial vehicles is becoming an increasingly demanding task: Shorter innovation cycles for high-quality and complex product solutions as well as an increasing number of model and equipment variants foreshadow the future challenges for manufacturers and suppliers. At the same time, customers' quality requirements for commercial vehicles are increasing, which in turn is reflected in their desire for high performance, low fuel consumption, and long service life and availability.
Increasing digitization in the commercial vehicle industry can help solve these challenges and provide manufacturers with a way out of the worsening requirements situation. Through the use of appropriate IT and software solutions, product use is becoming ever more closely integrated into the systems of the value chain. The efficient use and linking of the resulting vehicle data and existing process and defect knowledge is thus becoming one of the most important tools for ensuring that quality targets are achieved. The use of artificial intelligence as a solution approach in the manufacturing sector is thus associated with high economic potential.
Therefore, the newly launched research project "value chAIn" at the Laboratory for Machine Tools and Production Engineering WZL at RWTH Aachen University aims at developing a holistic, intelligent defect management along the value chain. The availability and performance as well as the early identification and sustainable elimination of process and product defects in production are to be optimized through the target-oriented feedback of production and field data and an uplifting, AI-based data analysis. In particular, the increase in transparency with regard to relevant dependencies between different instances of overarching value creation stages is targeted here. By means of the development and implementation of intelligent analysis methods for decision support, errors in the production processes are to be proactively eliminated, maintenance in the utilization phase of commercial vehicles is to be carried out with foresight, and the development of products is to be optimized.
This goal is achieved through the horizontal and vertical networking and evaluation of digital condition and fault information along the value chain of commercial vehicles. A key prerequisite is the cross-organizational provision of production and usage data. Artificial intelligence will be used to implement analyses on "Predictive Maintenance", for predicting and optimizing the maintenance of commercial vehicles and production units, "Predictive Quality", for predicting product quality in production, and "Process Optimization", for identifying optimal parameters. For this purpose, a decision support system is being developed that uses the results and data of the machine learning models as well as the knowledge of the employees to provide need-based information and derived actions for optimal decision selection.
The research project "value chAIn" will be carried out over the next three years in cooperation with the Laboratory for Machine Tools and Production Engineering (WZL) of RWTH Aachen University, the Fraunhofer Institute for Production Technology IPT and the companies KRONE Business Center GmbH & Co. KG (consortium leader), MAN Truck & Bus SE, i2solutions GmbH, DATAbility GmbH and IconPro GmbH.