What a $13.3 million line has to earn per hour
A contract packager received a state incentive award for a new automated production line: $13.3 million of capital, expected to create nine jobs. The record shows jobs created, not jobs removed, so the payback cannot come from labor saved. It has to come from what the line produces per scheduled hour. That number is not in the public record. This is it, worked from what is.
Two facts from the public record
- cited The state economic-development award record for the project: a building addition and a new automated production line, $13.3 million of capital investment, nine jobs created, six of them incented at a qualifying wage of about $30 an hour, and a state tax credit awarded against the investment. Event date 2026, verified September 2026.
- cited The plant's own pages: a careers page listing roles on second shift, third shift and nights, and an automation page stating it has installed several automated lines in the past few years. Both verified September 2026.
We do not name the company, its town or the agency; the figures come from the public award record and the company's own pages.
What the line has to generate, with every input labeled
The ledger is the analysis. Cited figures come from the public record and stay fixed; every estimate is a slider, and the totals re-run as you move them, so the plant's real hurdle rate, burden and shift plan replace ours in one motion.
Those are our assumptions where the record stops. Send your hurdle rate, burden and shift plan and we send back the version with your numbers in it: hass@hharesearch.org.
The record shows jobs created, not jobs removed, so the payback has to come from rate
Most automation cases close on labor: the cell replaces people, and the saved wages pay for the capital. This one runs the other way. The award record adds nine people to the plant at the same time as the line, and the record shows jobs created, not jobs removed, so the payback cannot come from labor saved. That is an inference from what the record reports, not a statement the record makes. On that reading the labor column is a cost on both sides of the ledger and the whole payback has to be found in what the line produces. Between about $430 an hour on three shifts and about $850 on two, for every hour the line is scheduled, held for five to seven years, and the shifts its own postings describe put it at the bottom of that range.
What a 10 percent rate shortfall costs
At the midpoint of that range, about $2.98 million a year, every single percentage point the line runs below its promised rate is worth about $29,800 a year. Ten percent below is about $298,000 a year, and about $1.79 million across six years, which is roughly 14 percent of the net capital at risk our estimate. Ten percent is not a disaster scenario. It is the ordinary distance between a rate that was quoted and a rate that gets delivered, and the quote is the seller's number: integrators bundle their own simulation into the sale. That work is competent, and it belongs to the party being paid. What does not exist anywhere in that transaction is a rate checked by someone who is not selling the cell our research.
The gap is measurable before the money is spent and verifiable after. In the published record, a simulation of a robotic tray-loading cell in food manufacturing predicted 4,110 packs and the installed line then delivered 4,110, at a 31.2 percent throughput gain; the same study's second case predicted 726 and delivered 723 cited. That is their work, not ours, and the second case is the more useful of the two, because it shows the method reporting its own small miss instead of rounding to a match. A plant that already has several automated lines in the ground has that many predicted rates and delivered rates that could be set side by side, and one more arriving.
What the plant receives
If they write, what goes back is a two-page note: the ledger above, every assumption stated so their team can swap in their real shift plan, case and pallet pattern, and the rate their integrator has quoted, plus what the line has to hold for the capital to close. Our standard is the one in the published record: the number is predicted before the plant spends and measured after. The full predicted-then-measured study is a paid engagement with a fixed scope; the note is not it. Replying carries no obligation.
What is cited, and what is ours
- The state economic-development award record for the project, event date 2026, verified September 2026. Supports the capital investment, the credit, the jobs created and the incented wage. Withheld.
- The award program's own rules on how the investment credit is earned and spread, verified September 2026. Supports the note that none of the credit offsets day one. Withheld.
- The plant's own careers page (shift structure) and automation page (several automated lines already installed), both verified September 2026. Withheld.
- Simulation study of a robotic tray-loading cell in food manufacturing, Foods 2024, 13(19):3121, published September 2024. Case A predicted 4,110 and delivered 4,110 at a 31.2 percent gain; case B predicted 726 and delivered 723, read from the full text. Their work, not ours.
- Burdened-labor multipliers and the competitive picture behind the unchecked-rate finding: HHA research, August 2026. Sources itemized on request.
Buying a line for rate?
Send what the line is meant to run. We reply with what it has to earn per scheduled hour for the capital to close, what a 10 percent miss costs, and whether it is worth thirty minutes on a call. We reply within two business days.