Geometry Assurance & Robust Design

The research group Geometry Assurance & Robust Design focus on the whole product realization process, from concept to production. The core competence in this group is within assembly modelling and analysis, robust locating system design, variation simulation, visualization, tolerance optimization, inspection preparation and use of inspection data for product and process control.

Geometry Assurance is the set of activities that aims to minimize the effect of geometrical variation in the final product. The Geometry Assurance & Robust Design research group works in all product realization phases:

In the concept phase the product and the production concepts are developed. In this phase, the concepts are optimized with respect to locator robustness and verified against an assumed production system by statistical tolerance analysis. The visual appearance of the product is optimized and product tolerances are allocated down to part level using variation simulation and sensitivity analysis.

In the pre-production phase the product and the production system are physically tested and verified. Adjustments are made to both the product and the production system to correct errors and prepare for full production. In this phase, fixtures and joining sequences are optimized with respect to geometrical quality. In this phase, inspection preparation and off-line programming of coordinate measurement machines and scanning equipment also take place. Inspection preparation and optimization of inspection points are supported by the use of statistical cluster analysis.

In the production phase all production process adjustments are completed and the product is in full production. Focus in this phase is on how to use inspection data to control production and to detect and correct errors. Lately simulation models, developed and used for tolerance analysis in the concept phase, have been fed with real time data to serve as digital twins. Here, selective assembly, automatic fixture adjustment and joining sequence optimization have been shown to reduce geometrical variation with up to 50%.