Last week I had a chance to speak with Andy Sherlock of ShapeSpace. My momentary interest was caught by a blog post – Clean up for PLM published on ShapeSpace blog. ShapeSpace is a small outfit trying out the water of 3D geometrical search for CAD and PLM. I found a problem of shape-based search quite interesting. Visualization of search can make search more useful. So, I decided to put some thoughts towards what happens in this space.
Google made efforts in the direction of going beyond text over the past year. The following product Google Googels available on Android devices brings a new experience in how you can find things visually.
ShapeSpace Part Browser
One of the core capabilities of ShapeSpace Part Browser is the ability to navigate and filter product shapes. In my view, it becomes useful for managing your part libraries. You can take a look on the following examples. Navigate on this link to see more videos.
Another solution I found in the same category is Bingo! from Sconce. I followed this company for some time and found they provided some very interesting examples of how to imply 3D search practices into product development. In the following video, you can see tight integration with Pro-E and Windchill. You can also get a brief about this solution from CIMData by navigating on this link.
I had to mention Geolus Component from Siemens PLM. The technology was acquired by Siemens PLM (then UGS) in 2006. You can see a historical article by Cadalyst: Geolus Search – The Google of 3D. You can find few examples of Geolus navigating your browser on the following link.
You can try to experiment with Geolus by accessing sample application via this link (registration required). You will see a pretty old web page with application with Siemens PLM Copyright (2008).
Why Search Stinks?
I think, visual is inspiring. However, visual and picture-based search are still not providing a level of reliability that can put it in a mainstream. Google Googels has limited capabilities that are very dependent on what type of picture you are trying to search. It works well on architecture, but fails on more casual examples. The same is in 3D… The precision of query definitions becomes a key driver of success.
(photo from Search Patterns book by Peter Morville)
What is my conclusion? Over the last 10 years Google converted “text based search” into a mainstream. Nevertheless, there is no one size search for every problem. 3D search is addressing an interesting problem of using geometry and shape to assist users in finding information. The result can be impressive. The query definition is still disappointing, in my view. I’d be interested to listen more about your Shape Search experience.