Self-learning application for determining the optimal settings of an industrial machine in a complex environment
Manufacturers are facing significant challenges in machine operations today. Complex production machines are very sensitive to changing production conditions and can produce non conform parts very quickly. For example, ambient temperature, air humidity and quality of raw materials from different suppliers have a strong influence on the quality of the production. Often these influencing factors can be measured throughout the production process and the quality of the produced part can be automatically qualified. However, not all production processes are yet fully automated and fine adjustment need to be made manually according to changes in the production conditions.
This requires a high level of expertise by the machine operator. With increasing numbers of influencing factors it is difficult for the operator to identify correlations and to react properly to changing production conditions.
With an aging work force, more and more experienced machine operators retire. Transferring all of the know-how will soon become a challenge.
Smart Machine Assistant was developed to support operators to find the right settings on their machines to produce quality conform parts under changing production conditions.
Smart Machine Assistant provides the optimal machine parameters to your operators
The solution Smart Machine Assistant is a combination of a standardized application and customized services to tailor the solution right to your needs.The application uses machine-learning capabilities to identify unknown relationships of machine parameters and KPIs. It gives machine operators concrete recommendations for adjustment of production parameters during operation to increase overall product quality and machine efficiency.
With Smart Machine Assistant you can optimize your production by having the right parameterization at hand. Intuitive visualisations help you to timely identify misconfigurations of the machine parameters.
The provided recommendations and available data in MindSphere support you in identifying the root causes of production drops.
In the first step the relevant assets from the shopfloor are connected to MindSphere and Smart Machine Assistant will be configured by MindSphere experts. After this Smart Machine Assistant will gather information about the production process and after a certain period of time the machine-learning component will start to make the first recommendations for the machine operators. From this point on, Smart Machine Assistant will continuously retrain the machine learning model and provide recommendations via the pre-built UI to the machine operator.
- Data normalization to create meaningful data sets of machine configuration, production information and IoT time series data
- Neural network based module that periodically creates proposal data sets and stores them in the IoT Value Plan
- Intuitive visualization of proposal data sets in a standard application
- Enabling industrial operators to improve machine and process productivity
- Improving the quality loop of production processes leading to reduction in scrap and increase in first yield pass
- Accelerating experiential learning process for operators leading to improvement in productivity and confidence in machine operations
- Enabling the fast and easy scaling to multiple machine types across multiple, global production networks