Step 1 – Use case framing

• Understanding of the industrial process, and definition of the metrics that will be targeted for optimisation
• Refinement of the end-user needs and definition of the web-app use cases and mock-ups
• Mapping of the key production parameters to be closely monitored and identification of the advanced ML methods to be implemented

Step 2 – Advanced method development

• Conception of a digital twin requiring the monitoring of more than 100 dynamic and static parameters
• Development of a customised ML method to identify and optimise the production process and enabling daily assessment of the production performance

Step 3 - Recommendation integration

• Replacement of faulty sensors that were identified thanks to the monitoring features
• Testing of the most promising production process enhancements such as input timing and quantity adjustments
• Quantitative validation of the enhancements for definitive integration