SPS 2019
Three questions for...Silke Lödige
Weidmüller offers a broad portfolio covering all aspects of digitalization and networking. At SPS in Nuremberg, the company will be demonstrating how customers can now generate added value from data. Silke Lödige, speaker for the trade press, spoke to us about machine learning and the future of AI applications.
What are the most important developments at Weidmüller since the last SPS and what will we be able to see in Nuremberg?
In recent years, Weidmüller has launched many communication-capable components for the provision and processing of process data. At SPS/IPC/Drives 2018, we exhibited a system for the first time that was completely controlled by our u-mation solutions, including the controller, remote I/O system, power supply, connection technology and signal converter. Our early positioning in the field of Industry 4.0, which we are now extending to future key market topics, is paying off here. We have already established a good basis here with our u-mation portfolio. At Smart Production Solution 2019, we will now be demonstrating how customers can generate added value from data - from smart sensors and connectors, industrial Ethernet components to machine learning solutions.
How do you get this added value from applications with big data?
The key questions for machine builders and operators are often: Is my machine's data good enough for me to generate relevant added value with the help of machine learning? Does my machine's sensor technology contain relevant information that is not being used to generate significant added value for me? Today, machine manufacturers and operators cannot answer these questions without an extensive project with a data scientist. This is why data analysis and modeling are still primarily carried out by data scientists. Today, their expert knowledge is mainly required to apply artificial intelligence or machine learning methods to the data and to develop models that recognize anomalies or predict errors, for example. The data scientist therefore works closely with the machine manufacturer or operator when developing models. Linking domain knowledge with machine learning methods is crucial.
How do you see this developing in the future?
Future success will largely depend on the extent to which companies succeed in integrating artificial intelligence into their products and services in order to tap into value creation potential. At the moment, the use of artificial intelligence and machine learning methods is reserved for a tiny group of data scientists compared to the number of engineers. This is slowing down the spread and broad industrial application of artificial intelligence. Technological approaches such as automated machine learning help to overcome this resource problem and promote the rapid penetration of industry with AI applications. We also want to drive this process forward with our software tool for the assisted generation of models.
Hall 9, Stand 351









