VIDEO: Programming Industrial Robot Toolpaths with 3D Scanning


Many manufacturing processes involve milling, drilling and machining, which require accurate G-Code toolpaths. These can be created by skilled craftsmen or by special software programs, but now it’s possible to scan existing 3D objects and create your toolpaths from there.

In the video above we look at how we can use such a solution to generate machine paths with a real-world setup. Without a CAD file, Darren Krasny, senior research scientist at Battelle, uses an off-the-shelf structured light sensor to scan a rough part spinning on a turntable.

“From that scan, we’ll able to interpret the data, recognize features on it, such as surfaces and edges and we’ll be able to create a raster coverage pattern that could be used for such things as surface polishing, deburring, et cetera,” said Krasny.

Using the associated software, Krasny can view the scanning process in action through a preview screen.

As the object rotates on the turntable, the scanner captures multiple images of the part and stitches them together to generate the 3D model, Krasny explained. With the model is completed, Krasny can interact with it through the software interface to plan tool paths along a selected surface.

“What our software does is it will fit a geometric primitive to that surface. So instead of dealing with tessellated data from the scanner, we can bring in an actual surface fitted to it,” Krasny explained. “Say we had a cylindrical surface, then we had a mathematical definition for it and now that we have that, we can generate a very smooth and accurate raster coverage pattern.”

The scan data is like CAD data and can be used to auto-generate a machine tool path.

“If I click the fill button, that locks in our pattern,” Krasny added. “On the screen display, we’ve got little markers that show you where the tool is going to be throughout the pattern and in the next step, the software will start to solve the generated path and find any necessary kinematic solutions, collision free, singularity free and joint limit free.”

For more information, watch the video above, get a copy of Krasny’s original presentation and visit the Battelle website.