Statistics
In mass production you cannot measure every part. You have to use statistical methods to make predictions about the actual geometric deviations of your parts.
Visualizations
In order to communicate analysis results in a team or with the customer / supplier you can sit together and look at a large pile of numbers and try to make sense of it. 😵💫
→ Or create some meaningful visualizations that convey the information contained in your data. With information you gain knowledge and with knowledge you can make reasonable decisions.
Automation
You can build statistical analyses and visualizations one by one in Excel or Minitab. 🥱
→ Or let them be created automatically. Get the latest reports with sophisticated plots directly in your inbox only seconds after the measurement data has been created.
Closed Loop
You have verified the product design by performing tolerance simulation with the tolerances on the drawings? Great 👍 And what will you do after the production of the parts has started?
→ Verify your assumptions (there are a lot and all of them are wrong 🫢): measure the parts, analyze the data and feed back the data into the simulation. This is a closed-loop tolerance simulation. This makes it possible, for example, to avoid tool corrections by identifying tolerance exceedances that are compatible with the part’s function.
Meta Models
Your simulation model can be very complex, so it takes a lot of time to run. Also it is not distributable to colleagues that don’t have access to the simulation software.
→ Reduce high-fidelity simulation models to fast-running meta models (surrogate models) that approximate system behavior with sufficient accuracy. We use traditional statistics and modern machine learning methods to create powerful meta models.
Our Services
- Statistical analysis of measurement or test data according to industrial standards.
- Reports, dashboards, internal apps built with R and Python. R is the programming language for statisticians and Python can do (almost) everything else.
- Closed-loop tolerance simulation as a service.
- Metamodeling as a service.
- Training and consulting for all these topics. Just tell us what you want to learn.