The calculation nobody does
Robot suppliers sell cycle time reduction. What they should be selling — and what you should actually calculate — is labour cost per good part, before and after. Everything else follows from that number.
Robot operating cost is small and usually overstated by suppliers: roughly 0.5–1.5% of machine value per year for maintenance and energy. The number that actually moves the answer is the reject rate, which we cover at the end.
A worked example with real numbers
A 300-ton machine producing a 180 g housing, three-shift operation, two operators per shift.
| Line | Manual | With robot |
|---|---|---|
| Machine cycle time | 32 s | 29 s (3 s faster, no manual take-out) |
| Operator handling per shot | 9 s | 2 s (loading only, no trimming) |
| Effective cycle | 41 s | 31 s |
| Output per hour | 87 parts | 116 parts |
| Operators required | 2 (handling is the constraint) | 1 |
Now the economics, using 3 shifts and a labour rate of $4.50/hour fully loaded:
Payback from labour alone: about 9.5 months. Add the output gain — if contribution margin per part is even $0.20, the extra 29 parts per hour across a working year is roughly $19,000 — and it drops under 6 months.
The finding that surprises most buyers
The cycle-time gain (3 s here, about 8%) matters much less than the labour-position gain (2 operators to 1). Almost all of the return comes from being able to run the machine with fewer people, not from making it faster. If your operation cannot actually remove a shift position when the robot arrives, the payback stretches well beyond three years and the calculation is not attractive.
What actually shortens the payback
- Multi-shift operation. The single biggest factor. A robot used one shift a day on a machine otherwise staffed three shifts saves nothing — it just reduces one operator's workload.
- Labour scarcity. In markets where you cannot hire operators at any wage, the calculation stops being about cost and becomes about whether the line runs at all. Then the robot is the only option regardless of payback.
- Overtime and weekend work. A robot that lets you clear a backlog without adding a shift turns a variable cost into a fixed one, which is worth more than the hourly maths suggests.
- Repeatability. Consistent pick-up removes variation from robot placement. On cosmetic parts this reduces scrap far more than most buyers expect when they build the business case.
Which robot type, and what it costs to automate
| Type | Best for | Typical cost | Limit to be aware of |
|---|---|---|---|
| 3-axis servo gantry (top entry) | Mould stays on the machine, pick and place, general purpose | Lowest, often $6,000–15,000 installed | Needs clearance above the tie bars; limited reach into deep moulds |
| 6-axis articulated | Complex geometry, in-mould work, insert loading, secondary operations | Highest, $25,000–60,000+ installed | Programming complexity; usually needs a specialist to set up |
| Cartesian / linear | Long thin parts, packaging end-of-line, high payload in a straight line | Moderate | Poor flexibility; rarely the answer for mould take-out |
| Servo press-mounted gripper | Very high volume, low complexity, single-cavity or simple multi-cavity | Low per unit, but fixed mould-position cost | Ties you to one mould; not upgradeable |
On 6-axis robots
A 6-axis robot is not "better" than a 3-axis gantry — it is a different tool. For simple pick-and-place from an open mould, a 3-axis servo gantry does the same job for a third of the price and with far less programming. Six-axis earns its cost when the motion genuinely requires it: insert loading, in-mould sensor work, or tending several machines. If nobody can name the specific reason, buy the gantry.
The costs people forget
Budget these alongside the robot price, because together they routinely add 20–40%:
- Mould changes. If the machine changes tools frequently, the robot has to be re-taught each time. This is often the largest hidden cost and it is a training issue as much as a hardware one — budget operator time for it, and buy robots with simplified teach pendants.
- End-of-arm tooling. Grippers are usually quoted separately. Vacuum cups, grippers and sprue cutters are all additional, and for a cosmetic part you need a proper gripper, not a vacuum cup.
- Safety guarding. Robot cells need guarding to be legally compliant. Often assumed to be included and frequently is not.
- Cycle verification during commissioning. The robot achieves its quoted cycle with the part it was demonstrated on. Yours will be slower until the motion profile is optimised. Allow a week.
The reject rate question
Run the calculation with a lower reject rate in the automated case, not a higher one. A robot that places consistently and takes the part without contact usually improves cosmetic quality, and any improvement in scrap is additional to the labour and output gains above.
Where robots do not pay
Honest cases: single-shift operation with no labour constraint; very short runs where programming time is not amortised; parts with inconsistent geometry that defeat fixed automation; and simply small enough jobs that a manual operator costs less than the robot's monthly depreciation. In those situations the machine still needs the automation — it is the business case that does not hold.
What to send us for a quotation
- Part drawing or photograph, with weight and dimensions
- Machine tonnage, and whether the mould is top-entry or needs side access
- Current cycle time and how many operators are on the line
- Shifts per day and annual volume
- How often the machine changes tools, and how many different moulds
- Your target — reduce headcount, increase output, or improve consistency. They lead to different solutions.
We will return the robot class, the installed price including end-of-arm tooling and guarding, the measured cycle time on your part rather than a demonstration part, and the payback calculation. If manual handling is the cheaper answer at your volume, we will tell you that too — it happens more often than the industry admits.
Common questions
Are injection molding robots worth the investment?
Usually yes on multi-shift operation, where the return comes from removing an operator position rather than from faster cycles. In a worked example — a 300-ton machine running a 180 g housing three shifts — a $28,000 installed robot saving one operator position at $4.50 per hour gives about $35,600 of annual labour saving, or roughly 9.5 months payback. Add the output gain and it drops under 6 months.
How long does an injection molding robot take to pay back?
Between 6 and 18 months in the cases where robots pay at all. The dominant factor is shifts per day, not the size of the output gain. A robot used one shift while the machine otherwise runs three shifts saves nothing, because you reduced one operator's workload rather than removed a position.
Should I buy a 6-axis robot or a 3-axis gantry?
For simple pick-and-place from an open mould, a 3-axis servo gantry does the same job for roughly a third of the price with far less programming. Six-axis earns its cost when the motion genuinely requires it: insert loading, in-mould sensor work, secondary operations, or tending several machines.
What are the hidden costs of installing a robot?
Budget an extra 20–40% on top of the robot price. The forgotten items are end-of-arm tooling (grippers and sprue cutters are usually quoted separately), safety guarding which is legally required and often assumed to be included when it is not, and cycle verification during commissioning. Mould changes are usually the largest — frequent tool changes mean frequent re-teaching.
Does a robot reduce the reject rate?
It can improve it, and any improvement is additional to the labour and output gains in the business case. Consistent placement and contact-free take-out usually improves cosmetic finish, because the variation comes out of the handling rather than the process. Run the calculation with a lower reject rate in the automated case.