Automated Scrap Prevention
Correlate production data to quality assurance results and learn which combinations of parameters lead to the least amount of scrap.
Getting more control over scrap by using machine learning techniques requires molders to go through a series of steps. Over time, this will lead to a more intelligent, data driven, production process. For a typical molder reducing the scrap rates down to < 0,5% from typical 2%-4% saves 200K to 400K per year.
Step 1: gather data from the production process
The first step in automated scrap prevention is to connect the manufacturing equipment, used during the injection molding process, to the MindSphere IIoT platform. This allows molders to gather valuable information directly from the production process.
Step 2: correlate production data to quality results
The production data can be correlated to results from the quality assurance process. This will provide insight into which production parameters have led to which results.
Step 3: learn good and bad parameter corridors
Once enough data is gathered, machine learning can be used to detect which combinations of settings across the production line lead to either good or bad manufacturing parameter corridors.
Step 4: predict and prevent scrap
After the machine learning step has been successfully completed, the trained model can be used to predict if a specific production setup will lead to an unacceptable amount of scrap. The application is able to propose alternative setups based on historical data.
Schedule a call!
Let’s have a quick call and see how this template can help your organization.
We’ll gladly give you a demonstration, facilitate a workshop or build an app in a day.