Freigeben über

The Science of Automation

  • Artikel

On a lark[1], I entered “industrial automation” in Wikipedia and founds some interesting links at the bottom of the page. Apparently, a few years ago, the IEEE[2] split their Transactions on Robotics and Automation into two publications in order to “bring more science to automation.” The new publication, Transactions on Automation Science and Engineering, is chaired by Professor Ken Goldberg[3] with Cal- Berkley.

The eminent Professor Goldberg has written quite a few papers on the subject of automation, but the one that stands out for me is the apply named Algorithm 93, or “Orienting Polygonal Parts without Sensors”. The algorithm, basically[4], measures the cross-sectional diameter of any 2-D polygon, analyzes the diameter versus angular orientation looking for minimums and maximums. After a sufficient amount of calculus incantations, out pops the result – the proper number of “squeezers” at the proper orientations to insure that all parts will have the same orientation on output regardless of their initial orientation.

Now, you could spend the time studying the paper, or do as I did and skip to the Java simulation. The simulation truly “makes the purpose and power of the algorithms come alive.” My challenge to the readers is – How many “squeezers” can you get from the simulation? You didn’t try hard enough if you get less than 6 <grin>.

- Jim
 

[1] Well, more out of desperation to get an article written on this subject.

[2] Disclosure: I am a member of the IEEE, haven’t been to a meeting since college.

[3] AFAIK, no relation to Rube Goldberg

[4] Weasel word for the fact I didn’t study read the paper.

Comments