LEDAS, a leading provider of software R&D (research and development) in the engineering software market, is pleased to announce a significant new customer for its in-house LEDAS Geometry Comparison (LGC) software component.
LGC is dedicated to detecting geometric differences between 3D models, and until now had only been marketed to and licensed by the CAD (computer-aided design) market. The customer will use LGC in its end-user PDM (product data management) software, part of its PLM (product lifecycle management) product line.
LEDAS first made contact with a new customer, a well-known engineering software vendor, several years ago while discussing our expertise in constraint solvers. Then a year ago, the customer began evaluating our trial version of LGC. It realized its PDM product could be enriched with functions that point out all the differences between two revisions of a 3D model, as well as all differences between two similar-looking 3D models.
After evaluating the quality and performance of LGC, the customer in October, 2023 decided to license LGC and integrate its code into their flagship PDM product. A confidentiality agreement prevents LEDAS from identifying the customer.
The Usefulness of Geometry Comparison
“We see many new areas for applying our LGC technology,” said Alexey Ershov, CEO of LEDAS. “We think there is a huge potential in the application of LGC to nearly every stage of the lifecycle of 3D models, especially with b-rep (boundary representation) models that lack history trees.”
Most PDM systems do not report specific geometric differences between revisions. But because LGC routinely deals with differences in revisions of the same model, PDM becomes an obvious market for it.
Another area for LGC is quality assurance by software that generates 3D models, like 3D kernels. LGC can be embedded in a testing environment to compare models generated during test runs against a reference sample.
LGC is also useful in comparing 3D assemblies. It provides quick searches for duplicates and similar-looking models in databases of 3D models. LGC even deals with cases in which identical/similar 3D parts have been translated or rotated with respect to one another.
How LEDAS 3D Geometry Comparison Works
As a technology, LGC is not specific to any particular 3D kernel. Implementations have been developed successfully for kernels by Spatial (ACIS), C3D Labs (C3D), and OpenCascade.
Comparing 3D models is a task that is more complicated than it looks, and so we use a custom implementation. In simpler implementations, other geometry checkers might use 3D Boolean subtraction. This approach, however, can produce bad output, because working with nearly identical 3D models is a worst-case scenario for 3D Boolean operations.
We have found that the performance of 3D Boolean operations is significantly worse than the performance of the custom algorithms we wrote for LGC.