What is the name of your product?

Automated Machine Learning Software!

What problem does your product solve?

Data analysis and model design are mainly carried out today by data scientists for analytics solutions Their expert knowledge is required to apply the methods of artificial intelligence to data and to develop models that recognise anomalies or predict errors, for example. Data scientists work closely with customers on model design. The drawback is that in future customers will want to do this directly themselves. In order to optimally integrate the domain knowledge of machine and process experts but at the same time automate model design steps, unsupervised and supervised machine learning have been intentionally combined. This combination helps create better models than would otherwise be achievable through a completely automated machine learning process. Initial applications of this type already exist in the areas of fintech, banking and marketing, but so far not in machinery and plant engineering.

What is the product’s main benefit for the customer?

With Weidmüller's solution, machine experts can in future help advance the development of models independently without having to be a data scientist. This ensures that the existing knowledge of processes and machinery stays within the company, as the engineers can update their domain knowledge themselves. The machine learning software enables machine or process experts themselves to tap into the performance of artificial intelligence. Weidmüller is therefore making the methods of artificial intelligence accessible to classic machine builders and operators. In future the lack of data scientists will no longer be such an obstacle to enabling artificial intelligence to penetrate the industry.

Why would you call your product “intelligent”?

Machine learning and artificial intelligence are the tools used for data analysis and interpretation. Platform-independent software covering automated machine learning significantly simplifies the application of machine learning and artificial learning and accelerates the production of analytics models.

Are there any other important points that you can think of?

The software solution is completely consistent, i.e. the software works end-to-end from model creation through deployment to the possibility of independent further development. The machine learning model can be directly deployed in the productive environment, even with an industry PC directly on the machine on which the software is running. Special attention has also been placed on user experience. This is crucial for the acceptance of artificial intelligence, as the analysis and model design steps must be comprehensible to users. It is important to reduce complexity for users, but at the same time ensure that they have the type of information to enable them to understand the impact of their knowledge on model quality.