PLI-Scale Volume: Can Your SMT Line Actually Take It?

Your SMT line's brochure speed is not its capacity. Capacity is what the line really places (rated speed, de-rated for real boards) multiplied by the share of staffed hours it actually spends placing. In one industry benchmark of contract-manufacturer lines, that share was only 54%. Before you buy a second line for a PLI-scale order, find out where the other 46% of your staffed hours go.

By the Mama Editorial Team. We scope camera projects for Indian factories. Last verified: 6 October 2026, against the PIB, IPC and CEERIS sources cited inline.

The volume is real. India's electronics production went from ₹1.9 lakh crore in 2014–15 to ₹11.3 lakh crore in 2024–25, and mobile-phone production alone reached ₹5.45 lakh crore. If you run a mid-size EMS or sub-assembly plant in Noida, Sriperumbudur or Pune, that growth shows up as a buyer asking: can you do 30,000 boards a month from next quarter? Saying yes from the machine brochure is how plants end up late.

Key takeaways

• The schemes are pushing volume to your door. The large-scale electronics PLI pays 4% to 6% on incremental sales over a base year. The component scheme (ECMS) expects production of ₹10,34,751 crore over six years, with PCBs and display and camera module sub-assemblies among its targets. That volume flows down the chain.
• Brochure CPH is not real CPH. A common practitioner rule of thumb is to de-rate a machine's IPC-9850 figure by 20–30% for real boards, and the IPC-9850 figure is already well below the headline "max CPH".
• Staffed is not placing. In the CEERIS benchmark, 72 contract-manufacturer SMT lines were staffed 112 hours a week but placed parts for only 63 of them: 54% line efficiency.
• The gap between 54% and 75% is often the size of the new order. In our worked example, closing it adds about 8,700 boards a month on the same line, the same two shifts and the same crew.
• A camera can show where the idle staffed hours go. It cannot tell you placement accuracy, mispicks or program efficiency. Your machine logs do that.

Step 1: Get a real placement rate, not a brochure one

Placement machines are sold on a headline "max CPH" (components per hour), which assumes ideal conditions. Hence a standard test. IPC-9850A places specified test components on specified test boards, "five different types of component parts with various quantities of each". The idea is a "truer comparison in placement speed and capability" across vendors.

Even IPC-9850 is optimistic for your boards. One SMT practitioner's explainer puts it plainly: most engineers "de-rate the manufacturers stated IPC9850 by 20-30% to estimate the real world CPH". Its example: a machine specified at 175,000 CPH with an IPC-9850 rating of 50,000, planned at 35,000. Treat the 20–30% as a rule of thumb, not part of the standard.

For capacity planning:

1. Take the IPC-9850 figure, not the max figure, for every placer on the line.
2. De-rate it 20–30% (the rule of thumb) for your real mix of parts. Then check it against what your machine logs report for your actual top-volume boards. The logs win.
3. The line rate is set by the slowest machine for that board, not by the total. If your high-speed chip shooter waits on the fine-pitch placer, the chip shooter's speed is irrelevant. (For the general version of bottleneck maths, see UPH, UPPH and line balancing.)
4. Divide by placements per board to get boards per placing hour.

That is the rate when the line is placing. The harder question is how often it is.

Step 2: Measure line efficiency the strict way

The most useful single number for SMT capacity comes from benchmarking firm CEERIS International. It defines line efficiency as the hours the volume pick-and-place actually places parts, divided by the hours the line is staffed. The definition is strict on purpose. Meals, breaks, partly manned lines and shift-change meetings all count as staffed hours, "even though the line may stop". In CEERIS's words, this way the ratio "does not allow manipulation".

What the benchmark found (same source):

Two cautions. The article is undated online and does not name the countries of its 19 factories: not an India survey, so a yardstick, not a target. And note the last row: across real factories a volume placer averaged 4,450 components per placing hour, a small fraction of any brochure number.

You do not need a consultant to get your own line efficiency. The CEERIS article notes that "the controller of most chipshooters records when they are placing parts and when they are idle". Pull a month of that log, divide by your staffed hours, and you have the number.

Step 3: Do the arithmetic before you sign

Here is a worked example. Every input marked "assume" is illustrative. Replace it with your own.

Assume: one SMT line with a combined IPC-9850 rating of 60,000 CPH, de-rated 25% to 45,000 real CPH for a 450-placement board. That gives 100 boards per placing hour. The plant runs 2 shifts × 8 hours × 26 days = 416 staffed hours a month. The buyer wants 30,000 boards a month.

At the de-rated rate, 30,000 boards need 300 placing hours, and you staff 416. On paper you have room. At the average efficiency you are a quarter short. The order fits only if you move from the 54% average to near world-class.

What the idle staffed hours cost you in wages. Assume a crew of 6 per line per shift: operators on the loader, printer, feeders, AOI and unloader. Pay them around the Uttar Pradesh Category I (Gautam Buddha Nagar and Ghaziabad) semi-skilled minimum: ₹15,059 a month (basic ₹14,173 + VDA ₹886), set as interim relief by the state's order No. 374/36-2-2026 of 17 April 2026, effective 1 April 2026. The VDA is revised periodically, so check your current rate. Assume a 1.3× loading for statutory contributions and leave.

The wage figure is not the real cost. The real cost is the 7,536 boards you cannot ship. Assume a conversion value-add of ₹50 a board (use your own quote). That shortfall is about ₹3.8 lakh a month of business you either turn away or deliver late, plus whatever penalty clause your buyer writes in. Late deliveries also cost you the next PO (why shipments slip).

A third shift or second line brings hiring, night-shift conditions or capex. Use the staffed hours you already pay for first. If you are already near 75% and still short, that trade-off is priced in second SMT line or a third shift.

Where the 46% usually goes

These are the usual categories of staffed-but-not-placing time on an SMT line. We give no percentages, because the honest split is the one your own floor produces in a two-week count.

Changeover deserves the most attention: a mid-size PLI-era plant often runs several part numbers for several customers. The fix is old and proven. SMED separates internal setup, which needs the line stopped, from external setup, which can be done while the last job is still running. Toyota used it to cut die changes "from hours to fifteen minutes by the 1960s". On an SMT line, external setup means pre-loading the next feeder trolley, pulling and checking the next stencil, and having the program verified before the last board of the current job leaves the printer. For a closer look at where EMS hours leak beyond the SMT zone, see where a mid-size EMS line loses hours.

What cameras can and cannot tell you here

What a camera can see. The idle hours your placer log can't explain. A camera over the printer and feeder-trolley area timestamps each changeover from last board out to first board in, and shows whether the next trolley was ready or being loaded while the line waited. One on the conveyor between reflow and AOI shows gaps in the board stream: the line was staffed but empty. One at the loader shows the minutes after shift start before the first magazine went in. Together these turn "54%" into named, timed causes.

What a camera cannot see. It cannot read placement speed, placement accuracy, mispicks, nozzle wear or tombstoned parts. Your placer logs, SPI and AOI do that, and they do it better. It cannot judge whether your placement program is well optimised. A camera does not replace the CEERIS calculation; it explains it.

People, not persons. A capacity study counts line events: stops, gaps, changeover minutes. It does not score individual operators. Tell your workforce what the cameras are for before they go live, using a proper DPDP-ready worker notice. Keep views away from customer artwork and screens. Many brand owners will raise this in an audit (see cameras in electronics/EMS plants).

Where Mama fits

Mama reads the cameras you already have, or a few we add over the printer, trolley area and conveyor. Each morning it sends you a short WhatsApp note on yesterday's line: how long it was staffed but not running boards, how long each changeover took, and which gaps came back again. It watches machines and flow, not people. If you want to know whether your line can take a bigger schedule, send a short phone video of your SMT floor. We will show you where the cameras should go and what they would measure.

FAQ

What line efficiency should an Indian SMT line plan on? Measure your own first: placing hours from the placer log divided by staffed hours. If you have no data, the CEERIS benchmark gives a sober starting point: 54% average for contract-manufacturer lines, 65% or more as a world-class factory average, about 75% for a single line. Plan on your measured number, not on the world-class one.

Should I use the max CPH or the IPC-9850 CPH for planning? The IPC-9850 figure, de-rated by the 20–30% rule of thumb for your real board mix, then checked against your machine's actual logged rate on your top-volume boards. The max CPH figure is a marketing ceiling, not a planning number.

Is my placement machine always the bottleneck? Often, not always. The printer, a slow fine-pitch placer, reflow conveyor speed or a full AOI queue can each set the rate for a given board. Check which machine is busy while the others wait.

When does a second line or third shift make sense? When your measured line efficiency is already high and the order still does not fit in staffed hours. In the worked example above, moving from 54% to 75% covers the gap on the same two shifts. If yours is already above 75% and you are still short, the extra-capacity decision is real.

Can I measure all this without cameras? Partly. Your placer log gives placing versus idle time, and a supervisor can time changeovers for two weeks. Cameras add the reason behind each idle block, daily.