Plain talk, with jokes

The second robot is easy. Here is why the first one is hard

October 1, 2026 · 3 min read

Shops that have one robot usually get a second within two years, and the second project takes half the time and a quarter of the anxiety. That is not because the robots got better. It is because the shop learned five things the first time that no vendor could have taught it. Here they are, so you can learn them before the first one instead of during.

The short answer

The first robot is hard because the shop is learning to be automated, not because the robot is complicated. Fixturing that was good enough for a person turns out to be inconsistent. Parts that a person picked from a heap turn out to need a tray. Nobody knew the real peak rate. Nobody owned the cell. And the first project was chosen because it was exciting rather than because it was easy. The Colorado molder in this site's casebook launched more than a dozen applications after its first cart-mounted arm, as the vendor published it. Nobody launches a dozen because the first was simple. They launch a dozen because the first taught them how.

Lesson one: fixtures are a discipline, not a part

A person loading a vise corrects for a part sitting slightly wrong without noticing. A robot puts the part where the program says and the vise closes on air. The first project forces a shop to standardize workholding across a part family so that every part locates the same way, and once that habit exists, every later cell inherits it. This is also why the second machine tending cell is quoted in days: the workholding question is already answered.

Lesson two: part presentation is half the project

Parts in a bin are a vision problem. Parts in a tray are a pick and place problem. Parts on a conveyor with a stop are a trivial problem. The first project teaches a shop to look at how parts arrive at every station and ask what it would take to make them arrive predictably, which is often a cheap fixture that helps the manual process too. The second project starts with that question already asked.

Lesson three: you did not know your peak rate

Almost nobody does until the first cell falls behind on a Thursday afternoon. The second project starts with a stopwatch, which is the single most valuable habit automation gives a plant and costs nothing.

Lesson four: a cell needs an owner

The first cell is everyone's and therefore no one's, until a jam at 2 AM settles the question. By the second cell the shop has an operator who runs robots, knows the integrator by first name, and has a card taped to the controller. That person is worth more than the second robot.

Lesson five: saying no

The first robot often goes on the hardest, most interesting job, because that is the one everyone talks about. It struggles, and the shop learns to put the second one on the boring repeat part with the long cycle, where it runs for years. The concierge on this site says no to the exciting job more often than people expect, for exactly this reason, and it is the piece of advice shops thank us for a year later.

How to learn them early

Pick the dullest repeating job with the longest cycle, standardize its fixturing before you buy anything, time its peak, name its owner, and write down what would make it a no. You have now done the first project's homework on a legal pad, and your first robot will behave like a second one. Luke Skywalker's first lightsaber duel went badly for a reason. The second one went better because of the first, not because the saber improved.

Describe your three candidate stations to Joe, the concierge on this site, and he will tell you which one is the boring repeat part that should go first. He is good at saying no to the exciting one.

Related on this site

• Your first robot should not be your hardest job
• Automating the wrong station
• What changed when a 54 person molder put one cobot on wheels
• Who will run the cell
• The ten application guides