Automate That

There are those that think that automation will be the salvation of American manufacturing. And there are those that think that automation is scary, replacing people with robots, costing American jobs.  While I’m not quite ready to accept as big a word as salvation, I do think that automation is the next big thing for manufacturing, period.


Automation cell (photo courtesy of Mazak Corporation)

Interestingly enough, the US actually lags in the installation of industrial  robots behind not only Japan but also China (source, International Federation of Robotics, ).   Which makes sense, because industrial automation is not just about labor reduction, it’s about repeatability and reproducibility and tolerances and quality.  And while there are plenty of massive  5-axis articulated arms, many industrial automation implementations take more mundane forms: tool changers, bar feeders and the like.

Which is not to say that automation is going to take over everything, despite what a few naysayers may be prophesizing.  For one thing, automated assembly lines need really high volumes of non-changing products to amortize the hefty upfront costs in equipment and programming. One long time user of industrial automation told me that they start thinking automation when you get over a million units a year volumes.  Of course it depends on the product size and complexity, but still if you have to amortize costs that can run 100s of thousands to millions of dollars, you need substantial volumes to make this make sense.

Humans are also still better than robots in many areas.  The Mark 1 A1 eyeball for example still can beat vision systems for high precision tasks that require active feedback (think threading a needle).  Flexibility is also huge for many product areas, and in general manufacturers are challenged to increasing deal with more flexibility and ability to change over quickly.

At Zebulon Solutions, we always look for the most cost effective path for volume production for our customers. Sometimes that does mean automation—we did a project using a Cartesian robot a few years back, and we’re supporting a medical device project on the DFx side that will need some serious automation.  We also count several robotics companies as customers (robots need to be productized too!).  But we’re not against either manual or semiautomatic solutions either. Frankly a good fixture costing a few thousand or even a few hundred dollars can often be the most cost effective solution.  But every project is different.


Leave a Comment