LEDAS Ltd :: Products :: Geometry Comparison :: Technology

LEDAS Geometry Comparison


Pinpointing Differences in 3D CAD Models

Comparing two similar-looking 3D models is like comparing text using tools like those provided with Microsoft Word. LEDAS Geometry Comparison (LGC) is the solution that pinpoints differences between similar geometric models, just as revision control systems do with text.

LGC uses tolerance control to separate important differences from those that are just small defects generated by import translations. When just part of a face is different (and the remainder of the face matches the face on the other body), then LGC subdivides the face and highlights the area that does not match. All the differences are collated and then presented to users in list and model views.

When one model is translated or rotated relative to the position of the other model, then LGC first computes the transformation to adjust the models in space before comparing them.

Mapping Faces Between Two 3D Models

After a model is modified, it still can look very similar to the original, but usually the mapping between faces is missing. For instance, if the original model is a component in an assembly, there is no way to replace it automatically with a modified model. This is because constraints in assemblies are applied to specific faces, and through the editing process the system loses the correspondence with faces of modified models.

LGC matches faces again between two models, and provides the result of the re-mapping in a report. It easily transfers the attributes of faces (such as colors, transparency, textures, and physical parameters) from the original model to the modified one; it recovers constraints, and it updates assemblies.

Robust and Fast Parallel Computations in the Cloud

The reason behind the development of LGC was to create a solution that would be much faster and more robust than existing software from competitors. The geometry comparison functions in mainstream CAD systems are based on non-stable, Boolean operations; models with just a thousand faces can take as long as an hour to process. In contrast, one-core LGC instance compares such models in seconds, and then delivers a report of all the differences it found and the faces it matched.

LGC was developed right from the beginning as parallel, multi-threaded, multi-process software for the cloud. This is a significant advantage for our technology. Our installation running on an Amazon cloud obtained exciting results: it was up to 5x-6x faster on 8 cores, and up to 15x faster on 32 cores.

   LEDAS & Partners News