Waiting for the Mechanic: What Sewing Machine Breakdowns Really Cost an Indian Garment Unit (₹)

When a sewing machine stops, the expensive part is usually not the repair. It is the minutes before the mechanic arrives, while the operator sits with her hands still and, once the bundles upstream of her run out, the rest of the line slows behind her. Price those minutes in ₹ per operator-minute and the choice between a faster call system, another mechanic or a spare machine becomes simple arithmetic.

By the Mama Editorial Team, for owners of Indian garment units of 200 to 1,000 operators. All ₹ figures are worked examples on stated assumptions, not benchmarks or any state's minimum wage.

Key takeaways

• A breakdown has two halves: waiting and repairing. Most registers log only "down 25 minutes", hiding the wait.
• One operator waiting costs little. A stalled line costs a lot. On our worked assumptions, an idle operator costs about ₹1.28 a minute. A 30-operator line that has run out of buffer costs about ₹38 a minute.
• Check the arithmetic before you hire. If your mechanics are busy less than about a fifth of the shift but operators still wait 10+ minutes, the problem is signalling and walking distance, not headcount.
• A pre-set spare machine is often the cheapest fix for critical operations: a 20-minute stop becomes a 6-minute swap.
• Cameras can time the wait, line by line, without naming anyone. They do not see why the machine failed.

Why "machine downtime" in your register is the wrong number

Prasanta Sarkar, an NIFT-trained textile engineer, makes this point on Online Clothing Study. If you start the downtime clock when the mechanic reaches the machine, you miss the time before. The breakdown actually began earlier, when the operator found the machine was not performing, so counting from the mechanic's arrival understates downtime. His fix is to have the operator note when the problem starts and when it is rectified.

In an Indian sewing hall, the gap between those two moments usually breaks down like this:

Only the repair row is the maintenance department's skill; the other rows belong to nobody, so nobody shrinks them. Our general piece on why breakdowns take so long to fix covers this five-step chain for machine shops. A sewing line differs in three ways that matter for your rupees:

1. Machines are cheap, people are many. Each machine has a paid operator tied to it.
2. A sewing line is a chain. If the stopped operation has no parallel machine, downstream stations run dry once the bundles in between are used up.
3. Spare machines are a real lever in sewing, unlike most machine shops.

A 2015 Apparel Resources workshop report on TPM noted that Indian companies "do not give enough importance" to losses from machine downtime, and argued for making the operator responsible for basic care of her own machine.

₹ per waiting minute: build it from your wage sheet, not ours

The cost of one waiting minute starts with what you pay per attended minute. That figure starts with your state's wage order.

Each state notifies minimum wages for the garment and tailoring trade by skill grade, and in most states the dearness-allowance part is revised twice a year, typically in April and October. The Code on Wages took effect on 21 November 2025 along with the other three Labour Codes. During the transition, existing state wage notifications broadly continue until the state notifies new ones.

In Karnataka, a big garment state, the government notified revised minimum wages on 22 May 2026. DLA Piper's summary describes a move from industry-specific rates to rates set largely by skill grade and zone. The Karnataka Employers' Association and others challenged it in the High Court (WP 16145/2026). LiveLaw reported on 21 September 2026 that the court declined interim relief without deciding the notification's validity. Use the notification currently in force for your trade, zone and skill grade, as your labour consultant reads it. We do not quote a state figure here because the rates and their legal status are in flux.

For the worked example, assume an all-in cost of ₹16,000 per operator per month. That covers wages at or above your state minimum plus the employer's PF and ESI. Then:

• ₹16,000 ÷ 26 working days ≈ ₹615 per day
• ₹615 ÷ 480 minutes ≈ ₹1.28 per attended operator-minute
• For a 30-operator line stalled at once: 30 × ₹1.28 ≈ ₹38 per line-minute

This wage-only figure is the floor of your loss. On a time-bound export order, lost pieces can mean overtime, which needs the worker's consent and is paid at twice the ordinary rate of wages (OSH Code s.27), or a missed shipment. And piece-rated operators bear part of the wait in their own earnings: a waiting operator is the one being let down, never the problem.

The worked example: three ways to spend on the same breakdowns

Here is one garment unit section of two lines of 30 operators each. All numbers are assumptions; replace them with yours.

• 6 breakdowns per line per shift on operations with no parallel machine
• Repair time 8 minutes once the mechanic is at the machine
• About 10 minutes of WIP buffer before downstream stations run dry; beyond that, assume the other 29 operators idle (an upper-end simplification)
• ₹1.28 per operator-minute, from above; 26 shifts a month

How to read it:

• Moving from A to B is worth about ₹84,000 a month across two lines. Whether that justifies a second mechanic depends on what one costs you, wage plus overhead, and on whether a cheaper signalling fix gets you to B first (see the next section).
• Line starvation, not the single operator, drives the cost. In scenario A, 290 of the 310 lost operator-minutes per event belong to people who did nothing wrong and whose machines are fine.
• A spare machine wins only on critical operations, and only if pre-set. A spare single-needle with the wrong folder, guide or stitch length is not a spare for a placket operation. The swap then takes as long as the repair.

Run the same logic with your own figures in the downtime cost calculator. For where breakdown sits against no-feeding, changeovers and bundle-wait, see where your sewing minutes go.

How many mechanics? The utilisation test before the ratio

Owners often ask for "the right mechanic-to-machine ratio". We could not find a published Indian benchmark worth quoting. Online Clothing Study's man-to-machine ratio counts maintenance staff inside total sewing-floor headcount, not as a separate ratio. Rules of thumb vary with machine mix and changeover load.

One documented case shows how far reality can drift. A 2012 Chulalongkorn University study of a Thai garment factory recorded about 790 breakdowns a month before its maintenance-system changes. The plant had two maintenance workers and 178 sewing machines: 68 on the line, 78 in reserve and 32 broken. After the changes, breakdowns averaged about 273 a month, and OEE rose from about 47% to 67%. It is not an Indian plant, but a pile of spare machines did not stop the breakdowns; maintenance discipline did.

Instead of a ratio, do this test with two weeks of your own data:

1. Calls per hour (λ) across the lines one mechanic covers. In our example: 12 calls in an 8-hour shift = 1.5 per hour.
2. Average repair time. In our example, 8 minutes.
3. Mechanic utilisation = calls per hour × repair minutes ÷ 60 = 1.5 × 8 ÷ 60 = 0.2, or 20%.

At 20% utilisation, a 12-minute wait is not a capacity problem. The mechanic is free most of the time. The minutes are going into discovery, signalling and walking: nobody saw the flag, the helper could not find him, or he was in the stores. Hiring a second mechanic will help a little. A call light, a fixed seat for the mechanic between lines, a small needle-and-parts kit at each line, and an operator trained to change her own needle will help more, and cost less.

As utilisation climbs, it changes. In standard queueing theory (M/M/1 queue), the average wait for a free mechanic is ρ/(1−ρ) times the average repair time, where ρ is utilisation. At ρ = 0.2 and 8-minute repairs that is about 2 minutes; at 0.8 it is 32. If your mechanics also handle line setting for frequent style changes, their real utilisation may be far higher than breakdowns alone suggest. That is when the second mechanic, or a dedicated changeover mechanic, pays.

Spare machines: rules that make them work

• Keep spares for operations with no parallel machine (collar run-stitch, placket, bartack, buttonhole), pre-set and labelled by operation.
• Swap first, repair off-line. Get the operator sewing again; repair away from the line.
• Count spares honestly. A "spare" awaiting a part for three weeks is a broken machine, not a reserve.

What cameras can show, in aggregate and without naming people

Existing security cameras over the sewing hall, read for the flow of work rather than for people, can produce these numbers daily:

• Station stopped while neighbours run: the start of an event, by line and position.
• Signal raised: a flag or light, if your lines use one, or a helper leaving the line.
• Help arrives: a person attends the stopped station. Where mechanics wear a distinct apron or coat colour, the arrival can be counted as a role, never as a named individual.
• Spare machine wheeled in versus repair in place.
• Restart: the station is sewing again, and the bundle pile upstream starts to shrink. On WIP economics, see bundles on the floor.

What an owner should receive is a line-level summary, not a video clip. For example: "Line 4, yesterday: 7 stops, median 11 minutes to help arriving, 3 stops long enough to starve the line, roughly 1,000 operator-minutes idle." There should be no per-operator ranking and no faces in the report. Footage of identifiable workers is personal data under the DPDP Act, so state the purpose (machine response measurement) in a worker CCTV notice before you start.

What cameras cannot tell you

• Why the machine failed. A camera cannot see hook timing, a worn feed dog or bad thread. That needs the mechanic's log.
• Micro-stops the operator fixes herself, such as re-threading or a quick needle change, unless the view is close and the event is clear. Short events will be under-counted.
• Hidden stations. Pillars and hanging bundles block some machine faces.
• Lost output in ₹. The camera gives minutes; your wage sheet and order book turn them into rupees.

Treat camera timings as estimates good enough to rank lines and shifts, not as stopwatch-exact records.

Where Mama fits

Mama reads the cameras a garment unit already has and sends the owner a short note each morning: which lines stopped, how long they waited for help, and roughly what that cost. It watches machines and flows, counts people only in aggregate, and never ranks operators. If you want to know whether your existing views cover your critical operations: Leave an email or phone in the form below, and Mama will get back to you with a camera plan.

FAQ

What is a reasonable mechanic response time on a sewing line? No published Indian standard exists that we could verify. Measure your own median time from stop to arrival for two weeks, then work to shrink it. If your mechanics are lightly loaded but waits exceed 10 minutes, look at signalling and walking distance before headcount.

Is the operator paid while she waits for the mechanic? A time-rated operator is paid for the shift, so the wait is pure cost to you. A piece-rated operator may lose earnings for the wait. Check how your state's wage rules treat idle time for piece-rated workers with your labour consultant.

How many spare machines should a garment unit keep? Base it on critical operations, not a percentage of machines: one pre-set spare per frequently failing operation with no parallel machine, reviewed monthly against actual use.

Can cameras replace the maintenance register? No. Cameras time the stop, arrival and restart; the register records the fault and parts. You need both.

Will operators see this as surveillance? They might, if it is reported per person. Report by line and shift, tell workers what is measured and why, and use the numbers to fix signalling and spares. Do not use them to discipline the person who was left waiting.