Generic AI for virtual geometric validation
Virtual geometric validation (VGA) – also known as geometry check, geocheck or collision analysis – accompanies the entire process of digital product development. Its aim is to increase the maturity level of digital prototypes and improve data quality in order to ensure a "buildable" product at the end.
In the VGA process, geometric conflicts such as collisions, minimum distances or distance violations are recognised, evaluated and communicated on the basis of the digital prototypes (CAD data) so that potential problem areas can be rectified at an early stage.
Status quo: large amounts of data, growing lack of transparency and high manual effort
The technology is already in daily use: millions of component pairings from OEMs in the automotive, aerospace and commercial vehicle industries are analysed fully automatically and geometric anomalies are immediately available. For customers, this means significantly lower costs, less manual inspection work and a transparent, accelerated development process.
Thanks to the pre-trained model, the generic AI can be used immediately – without lengthy training phases or high implementation costs.
Solution: faster process through the use of tried-and-tested AI
Without intelligent technological support, the manual effort is enormous: experts have to check every possible problem area and assess its criticality. To do this, the point of conflict must be visualised, inspected – usually using a 2D section or in transparency mode – and then documented using a manual screenshot.
For large products with often hundreds of thousands of conflict points, this quickly leads to enormous costs, high time expenditure and delays in the entire development process.
Generic AI accelerates the entire VGA process and makes the digital validation of virtual prototypes significantly more efficient. Integrated into quality monitor or customer systems, the AI takes over central work steps – fully automated, reliable and traceable.
- Fully automated data preparation: The pre-trained model structures the data and enriches it with relevant additional information.
- Intelligent image generation: The AI automatically generates an image for each potential point of conflict that clearly and unambiguously depicts the situation. Four standardised viewpoints enable experts to make a quick and reliable assessment – without having to do any visualisation work themselves.
- Specific assessment suggestion: The AI provides a percentage suggestion for the criticality of each point of conflict – including an assessment of how confident it is.
Result: With the image and assessment suggestion, the expert can make a decision immediately. This significantly reduces manual effort and makes the VGA process 50% more efficient.
Successfully productive in use
Virtual geometric validation (VGA) – also known as geometry check, geocheck or collision analysis – accompanies the entire process of digital product development. Its aim is to increase the maturity level of digital prototypes and improve data quality in order to ensure a "buildable" product at the end.
In the VGA process, geometric conflicts such as collisions, minimum distances or distance violations are recognised, evaluated and communicated on the basis of the digital prototypes (CAD data) so that potential problem areas can be rectified at an early stage.
