Vision guidance for depalletization

If X = 1000 Y where X = pictures and Y = words, what would be the equivalent expression for a movie? One movie is worth a hundred thousand words?

That’s what “Robotic controls double throughput” in the July 2011 Control Engineering magazine left me wondering.

The article describes a vision guided robotics application where a 3D vision system (it uses 2 Sick Ranger cameras,) enables a robot to unload a pallet of large sacks of coffee beans. The problem I had was, if you know anything about laser triangulation for 3D you’ll know the laser line has to scan over the surface. In other words, there has to be relative motion. But the article was not very clear how this was actually achieved.

So I went to the integrator’s website – a North Carolina company by the name of Concept Systems - and found a link to this marvelous movie of the system in operation. For your edification and delight I shall embed it below.

Now, if you’ve watched the movie – a very professional production devoid of the ghastly plinkety-plink music sometimes used as a soundtrack – you’ll know exactly how the relative motion is provided.