Physna, an industry leader in ‘Geometric Deep Learning’ technology, has recently launched a geometric search engine named Thangs. Physna is calling Thangs the 3D world’s Google x GitHub crossover, i.e., it is supposed to be that powerful platform for 3D models.
Thangs uses Deep Learning search algorithms that focus on the polygons or triangles, making up the 3D model’s volumes to be indexed. Also, ‘version control functionality’ and ‘compatible part predictions’ are plus points for the 3D community. Thangs 3D model search engine launches with more than a million searchable objects and aims to be a ‘3D Google’.
Engineers, industrial designers, and 3D-printing enthusiasts face major challenges of working with 3D data. As it is easy to search for 2D data like text or image in Google, but dealing with 3D data needs another platform, which Thangs provides.
Thangs is the first open product of Physna, and it is believed that the industry is trying to democratize some of the technologies at the core of its enterprise products. It is easy-to-use for a layman, and powerful enough for a leading aerospace CAD engineer, as claimed by Paul Powers, CEO of Physna.
How does it work?
- Users may upload parts.
They get the suggestions regarding the usage, compatible components with the uploaded part, and where-to-use those parts.
- A text-based search box is also available to find the desired parts.
The search can be based on the object’s physical properties, measurements, and features. And some relevant suggestions about the part’s function, cost, materials, and performance are also provided.
- Version control is automated, much like Github.
Users are free to ‘like’ (by which they can save the model for later use) or ‘comment’ the models. Thus, it also introduces collaboration tools for 3D model creators.