In automation charter meetings, one station type gets named with suspicious frequency: near the visitor path, single-motion, loose takt, a stop that does not block the main line. None of those traits is wrong by itself. The trouble is they rise to the top of the list because they are easy to present—not because they pay back fastest on the ledger.
The first cobot in an SME often carries two jobs: remove repeat labor at one station, and prove to the board that automation is viable. Those jobs diverge. Evidence projects want stability, speed, and a story non-specialists can grasp in one walk-through. Capacity projects want the bottleneck, peak season, and a cell that actually carries load when hiring hurts. If station one mainly serves the first job, the slide deck wins and the line “does not feel it.” Every later conversation about ROI, replication, and the next PO starts from a thin sample.
Risk minimization is the first reason the wrong station goes first
Touch the bottleneck and delivery dates, customer promises, and piece-rate payroll all move. Touch the showcase station and—even with two extra weeks of tuning—the external story stays “we are exploring.” Visibility is the second driver: cobot investments need internal sponsors; the cell guests see becomes the company video. A good shot is not capacity. The third layer is technical avoidance: the hardest stations hide messy incoming parts, loose fixtures, frequent changeover. Commissioning on a clean station goes smoothly; it does not prove the dirty one will.
The bill often arrives in year two, not year one
When the board asks for more units, finance models payback from station one. If that station was never chosen to save labor hours—light repeat work, or cheap manual labor—the cycle looks long and automation loses the budget call. That is not always arm price; it is the wrong sample station. Replication hurts the same way: gripper strategy, takt breakdown, safety argument templates learned on an easy cell get rebuilt on the hard one. “Same as station one” only bites when station one never represented your real process difficulty.
How to tell if station one stood in the right place
You do not have to start with the nastiest job in the plant. But run an honest filter: in the last twelve months, did overtime, hiring pain, or scrap variance from repeat work at this station rank in the top three? If yes, it deserves the money. If no, admit station one is mainly learning curve or internal proof—and lower the ROI promise and replication timeline accordingly. Two hard checks rarely lie: was worst-case on the proof list, and would you default peak-week output to this cell? If both are no while station two is already in discussion, station one’s job was probably mislabeled.
We are often station one—we understand that role. The point is not “always pick the hardest.” It is “tell the truth about why.” Capacity and proof are both legitimate; do not use a proof project’s story to approve a capacity budget. Putting worst-case in the charter matters more than putting the arm by the door. Choosing the wrong first station is not shameful. Using the wrong sample to plan the whole factory’s automation future is.
