Automated Machine Learning (Industrial AutoML)

Create and use end-to-end machine learning solutions faster. You only need your application knowledge!

With Weidmüller Industrial AutoML you can optimize operations, increase product quality and create new business models by taking advantage of advanced analytics. As a machine or process expert you can easily build, deploy and operate machine learning models without needing any data science skills. The AutoML tool empowers you to transfer your data and domain knowledge into ML models that generate value for your business. Models can be deployed in existing manufacturing environments to put real time analytics and insights at the fingertips of production workers and stakeholders across your organization.

Download free whitepaper "Machine Learning automation for industrial applications" here

Your benefits at a glance

Accelerate innovation

Benefit from advanced analytics while using your existing machine data and domain knowledge. No extra training required. Build your own machine learning models within minutes

End-to-End solution

Create, deploy - on-premise or in the cloud - and operate machine learning models. You can continuously improve model performance easily by retraining models as you gain more insights and collect more data from machines and processes.

Build customer relationships and new business models

Increase customer satisfaction with improved products and services, and achieve a better understanding of your customer’s needs and behavior powered by machine learning technology developed by you.

Benefit from machine learning without data science skills

Watch our 2-minute animation for a compact overview of the Weidmüller Industrial AutoML solution including its key benefits and an explanation on how it works.

Product modules and features

Create, validate and export your own machine learning models with ModelBuilder

This is based on the following step by step approach:

Import and explore your machine and process data
Assess your data based on automatically generated quality indicators (e.g. missing values)
Enrich your data by creating custom features
Contextualize your data e.g. by defining what's normal behavior and what isn't
Choose the kind of machine learning model to create e.g. anomaly detection or classification
The tool then automates the model creation process, including feature engineering, required preprocessing and post processing operations
Select the most appropriate model from the models created, based on criteria such as model performance and plausibility

Deploy, configure and score your own machine learning models with the ModelRuntime

Deploy your models where you need them: on-premise or in the cloud
Connect your machine to ModelRuntime by configuring data sources e.g. a database
Import the created models and assign them to the specific machine. Multiple models can be used for the same machine
Schedule the model execution according to your requirements
Visualize results using the embedded GUI
Use and configure ModelRuntime programmatically by using the provided interfaces
Put model results into actions by importing model outputs into the existing machine or manufacturing systems

Learn more about Weidmüller Industrial AutoML from our experts

Industrial AutoML enables domain experts to create and optimize machine learning models

Learn more about why machine learning is transforming the industry, how Industrial AutoML is accelerating this process and how you can benefit from it by speaking with our responsible product manager Dr. Carlos Paiz Gatica.

Industrial AutoML opens up the complex world of Data Science to you

Learn more about the data science background of Weidmüller Industrial AutoML by our Data Scientist Dr. Daniel Kress

Learn more about Weidmüller Industrial AutoML in use

Boge Compressors
GEA Dairy Industry
MULTIVAC packaging
Weidmüller Galvanic

Boge Compressors

APPLICATION AREA

Prediction of the wear behaviour of critical compressor modules

Monitoring of compressed air availability

BENEFITS

Delivery of the relevant sensor data

Optimised production processes and test procedures

The Automated Machine Learning Tool helped me to create my own analytics models in a short time without having any data science know-how. I was positively surprised about the good results the tool produced based on my application knowledge about the compressor. The model creation process and model selection was intuitive and easy for me to follow.

Dr. rer. nat. Christian Heesing, BOGE

GEA

APPLICATION AREA

Automatic detection of anomalies in the behaviour of separators in the dairy industry

Scope: 500 integrated machines

BENEFITS

Increase in efficiency and productivity

Transfer of application knowledge into an algorithm

Extension and development of offered services

More information

We were fascinated by the solution, as we have a lot of process engineers who are very familiar with the machines and who are, to a certain extent, able to interpret the data. With Weidmüller's help, we can now transfer this knowledge to an algorithm

Matthias Heinrich, Manager Digital Solutions, GEA

With the help of Weidmüller's AutoML software, we were able to generate an initial model for detecting anomalies with fairly little effort. This has already identified 97 per cent of the anomalies in the actual process. We especially like how easy the software is to use. The ability to mark normal and abnormal time ranges for model building is very well implemented.

Dr. Martin Roth, Data Scientist, MULTIVAC

Der AutoML Modelbuilder ermöglicht eine einfache und schnelle Erstellung sowie Validierung von Verfahren zur Anomaliedetektion auf Basis historischer Daten von Motorströmen eines Fräsprozesses. Die Bedienung ist intuitiv und unterstützt sowohl Anfänger als auch erfahrene Data Science Anwender effektiv.

Andre Schmidt, Automatisierung / Softwareentwicklung, G.Kraft Maschinenbau GmbH

Weidmüller Galvanic

APPLICATION AREA

Monitoring of pumps in a galvanic plant

Sensors collect data from pumps e.g. vibrations, electrical current, temperature or flow rate
Target is to detect unusual pump behavior e.g. a bearing damage to avoid unplanned downtimes and ensure process quality

BENEFITS

Easy and quick testing of predictive monitoring of pumps

Speed: a model is generated in less than 1 hour with ModelBuilder
Intuitive usage: you are taken step by step through the process of model generation
Flexible on-premise deployment with ModelRuntime