Production machines are subjected to heavy loads and run around the clock at optimum capacity. The manufacturing industry cannot afford machine downtime or high scrap rates if it wants to remain competitive on the international market. Processes often run at high machine speeds and with low manufacturing tolerances. To achieve efficient production at high product quality, process and machine conditions must be monitored continuously. This process monitoring is done with the help of sensors, which must be placed close to the manufacturing process for maximum information gathering. However, conventional sensor systems are usually cable-based and are not suitable for machine processes with complex motion sequences. Also production processes must be monitored using real-time data in order to be able to react in good time to deviations in the process. In the "WiMuSens" research project, the Fraunhofer IPT and its research partners are developing a wireless sensor platform that transmits real-time signal data at high speed, evaluates it using artificial intelligence and autonomously optimises machine control.
The quality of the manufacturing process is determined by various process parameters such as spindle speed and feed rate. Wear can thus lead to the machine and process parameters no longer optimally matching the manufacturing process and to a drop in production quality. In the "WiMuSens" research project, the Fraunhofer IPT is therefore working on a wireless multi-sensor system that can detect quality changes in the infra- and ultrasound range using sensors with different frequency bands. New technologies such as 5G or Bluethooth 5 make it possible to transmit fast even multidimensional data sets and evaluate them using suitable algorithms. After data pre-processing, in which relevant information is filtered out, the data set is compressed so that the amount of data is reduced while the information content remains the same. For the rapid evaluation of the machine data, the Fraunhofer IPT uses analytical and stochastic models in the project. Self-learning algorithms from the field of artificial intelligence are used to capture all relevant process information and provide it to the machine control system in real time.
The task of the Fraunhofer IPT in the "WiMuSens" project is the development of software components for data pre-processing, process monitoring and process optimization based on artificial intelligence. In addition, the Fraunhofer IPT is responsible for the machine connection to the sensor platform and the validation of the process. The wireless sensor platform can be individually configured through the use of hardware and software modules and thus offers enormous potential for the stable process control and monitoring of production machines.