# Sewing Floor Management: What a Supervisor Actually Controls, and How You Know > Most sewing floor management advice lists what to control — line balancing, WIP, hourly output, needle policy — without saying how a supervisor knows any of it is happening. This covers both: what you control, what you check, and what tells you it is slipping. **Source:** [https://scanerp.pro/blog/sewing-floor-management-garment-factory.html](https://scanerp.pro/blog/sewing-floor-management-garment-factory.html) --- # Sewing Floor Management: What a Supervisor Actually Controls, and How You Know Santosh Rijal · September 16, 2026 · 9 min read · Production **TL;DR — Direct Answer:** A sewing floor supervisor controls seven things: **line balance, WIP on the floor, hourly output against target, in-line quality, needle policy, machine condition, and aisle space**. Every guide lists those. The part usually missing is how the supervisor *knows* any of them is slipping before end of day — which is a data-collection problem, not a discipline problem. For each control area below: what you control, what you check, and what tells you it is going wrong. And an honest closing point — finding out faster does not fix anything by itself, it just buys back the hours you would otherwise lose. Search for sewing floor management and you get a competent list: balance the line, control WIP, track hourly output, run in-line QC, keep a needle policy, service the machines, keep the aisles clear. I have no argument with the list. I have run a sewing floor for years and it is the right list. The gap is that almost nobody says how the supervisor finds out. One widely-repeated piece of advice is to "record hourly output per operator or line" — with no word on how that is done across several hundred machines by one person who is also meant to be solving problems. On a small floor you count. Past a certain size, counting becomes the job, and the problems stop getting solved. So this article takes the same control areas and answers three questions about each: what the supervisor actually controls, what they check, and what tells them it is slipping. ## The Supervisor's Real Job Is Removing Obstructions Start here, because it determines everything else. A sewing supervisor is not a recorder. The value of the role is in the twenty minutes after something goes wrong: a machine down, an operator absent, a bundle short, a component that has not arrived from cutting. Every minute a line spends waiting for the supervisor to notice is production you do not get back, because a sewing line cannot make up an idle hour later in the day. That means any system, paper or digital, that makes the supervisor spend more time recording and less time walking is working against you. This is the test I apply to anything we add on our floor: does this take the supervisor off the floor, or does it point them at the right machine faster? Most CMT floors run one supervisor per line or per two lines, plus a floor in-charge above them. The higher the ratio of machines to supervisors, the more the supervisor depends on being told where the problem is rather than finding it by walking. ## Line Balancing: Seeing a Bottleneck Before It Is Obvious **What you control:** who sits where, which operations are shared between two operators, and where you add a machine. Line balancing is the technique of keeping each operation's capacity close to the line's target so that no single operation throttles the rest. [Online Clothing Study's definition](https://www.onlineclothingstudy.com/2013/02/what-is-line-balancing.html) is the standard one: the operation with the lowest hourly production sets the output of the whole line. **What you check:** the operation-wise capacity against the target. Classically this is a capacity study — stopwatch, five cycles, an hourly figure per operator — plotted as a pitch diagram so the short operation stands out. **What tells you it is slipping:** here is the honest problem with the stopwatch method. It is a snapshot. You balance the line on Monday morning with good data, and by Wednesday an operator has been moved, a machine has developed a fault, and the balance is gone. Nobody re-runs a full capacity study mid-lot. The symptom everyone recognises is a pile of work in front of one machine and idle hands after it. But a pile takes hours to build, and by the time it is visible from across the floor the hours are spent. The earlier signal is the rate of change: work arriving at an operation faster than it leaves, sustained over two or three hours. That is a comparison of counts in and counts out per operation, which means bottleneck detection and output tracking are the same measurement problem wearing two hats. [The standard remedies](https://www.onlineclothingstudy.com/2012/01/7-ways-to-reduce-wip-from-bottleneck.html) — method improvement, work sharing, an extra operator or machine — are all available to a supervisor who knows by 11 AM instead of at 6 PM. If you want the arithmetic behind the target figure itself, we have separate guides to [sewing line efficiency calculation](/blog/sewing-line-efficiency-calculation-garment-factory.html) and [SAM and SMV calculation](/blog/sam-smv-calculation-garment-industry.html). ## WIP Control: Bins, Trolleys, and Money Sitting on the Floor **What you control:** how much cut work is released onto the floor, and how it is held between operations. Bins and trolleys are not housekeeping. A defined container per operation caps how much can queue there, which is a physical limit on WIP that needs no software at all. **What you check:** bundles between operations, and how long the oldest one has been sitting. The second half of that sentence is the one factories skip. **What tells you it is slipping:** age, not volume. A trolley with thirty bundles that turns over every hour is fine. A trolley with twelve bundles where three have been there since yesterday is a problem, and the two look identical from a distance. On our floor the supervisor's lot view flags anything that has been stalled beyond six hours, which catches the bundle that quietly fell behind a machine as well as the one waiting on a missing component. Why it matters in cash terms: WIP is fabric you have paid for, cut, and partly sewn, which cannot be invoiced until it leaves the building as a garment. It also lengthens the gap between making a mistake and finding out about it — a defect sewn on Monday and discovered at checking on Thursday has had three days to repeat. That delay is usually the bigger cost. I have written the full argument up separately in [WIP tracking in a garment factory](/blog/wip-tracking-garment-factory.html). ## Hourly Output vs Target: The Collection Problem **What you control:** the target itself, and the response when the line falls behind it. **What you check:** pieces produced this hour against pieces planned this hour, per line, and ideally per operation. **What tells you it is slipping:** this is the one where standard advice runs out. Recording hourly output per operator sounds reasonable until you try it on a few hundred machines. Someone has to walk the line, count bundles, write the figures down, and total them, every hour, and that person cannot simultaneously be fixing the thing the numbers reveal. So in most factories the hourly board is either filled in for one or two showcase lines, or filled in retrospectively at the end of the day from memory, which makes it a record rather than a control. The way out is to stop treating counting as a separate activity. If the handover of work is already a recorded event, the hourly figure assembles itself. On our floor each bundle carries a QR code holding lot, article, colour, size, component and quantity, and every scan records operator, machine, operation and timestamp. Nobody writes a tally sheet, because the tally is a by-product of work that was going to happen anyway. Our bottleneck summaries then go out on WhatsApp at 11:00 and 15:00, which is early enough in each half of the day to still do something about them. There is a second benefit that is easy to miss. Because each scan is tied to an operation and its predecessors, work is held until the components it needs are complete — dependency gating. An operator does not start a joining operation and then discover half the collars are still upstream. Stuck operations are swept every two minutes, with a five-minute Cloud Function backstop, so a scan that half-registered does not sit invisible for the rest of the shift. That is the real lesson from our floor: most of the value was not the dashboard, it was removing the ways work can silently stall. The mechanics are in our write-up of the [bundle system for production tracking](/blog/bundle-system-garment-production-tracking.html). ## In-Line QC and When to Stop a Line **What you control:** where the checking points sit, and the authority to stop work when a fault repeats. **What you check:** a sample at the end of critical operations, rather than only at the end of the line. The whole reason for in-line checking is to catch a fault while the number of affected pieces is still small. **What tells you it is slipping:** repetition on the same operation. One skipped stitch is a piece. Four skipped stitches from the same machine in an hour is a machine problem, and every piece that machine has produced since the first one needs looking at. The judgement call is when to stop the line, and it is genuinely a judgement — stopping costs output, not stopping costs rework at a worse rate. What makes the call easier is knowing how many pieces are exposed, which is a WIP question again. Most floors run end-of-line checking as the formal gate and in-line checking as the early warning. The defect data itself is worth structuring properly; we cover that in [what floor QC software needs to do](/blog/garment-qc-software.html) and [what a good DHU rate looks like](/blog/dhu-benchmarks-garment-factory-2026.html). ## Needle Policy: The Control Buyers Audit Hardest **What you control:** the broken-needle procedure and whether it is followed when nobody is watching. **What you check:** the log. The standard [broken needle policy](https://www.onlineclothingstudy.com/2019/11/broken-needle-policy-for-garment.html) requires that a broken needle is reported immediately, all fragments are recovered and taped to a log sheet against the issued replacement, and the affected garment is isolated until the pieces are accounted for. Records are typically kept for at least a year for audit purposes, and factories are expected to follow the specific policy their buyer issues. **What tells you it is slipping:** a log with gaps, fragments that do not add up to a whole needle, or a replacement issued without a broken one surrendered. Buyers audit this harder than almost anything else on the floor, because a metal fragment in a shipped garment is a recall and a reputational event, not a quality deduction. This is one area where paper discipline genuinely works, provided someone checks the paper. ## Machine Maintenance: Cheap When Scheduled **What you control:** the cleaning and oiling schedule, spare part availability, and how fast a mechanic reaches a stopped machine. **What you check:** daily operator-level cleaning, and periodic preventive servicing rather than repair-on-failure. **What tells you it is slipping:** the same machine number appearing repeatedly in breakdown reports, and rising mechanic response time. A machine down for forty minutes because the mechanic was at the other end of the floor is forty minutes of one operator's output plus whatever the downstream operations lost waiting. Tracking which machine stopped and for how long turns a vague sense that "that machine gives trouble" into a replacement decision. We go into the detail in our guide to [machine maintenance tracking in a garment factory](/blog/cmms-machine-maintenance-garment-factory.html). ## Floor Hygiene and Aisle Space **What you control:** the layout, and the standing rule about what may sit in an aisle. **What you check:** that trolleys can pass, that fire exits are clear, that thread waste and cut fabric are cleared during the shift rather than at the end of it. **What tells you it is slipping:** bundles on the floor instead of in containers. It is the most reliable single indicator on a sewing floor, because it means WIP has outgrown its containers and nobody has dealt with it. It also compounds: work on the floor gets stepped on, soiled, and lost, and lost bundles are the hardest kind of short to explain to a buyer. Layout decisions that make this easier or harder are covered in [garment factory layout and sewing line setup](/blog/garment-factory-layout-sewing-line-setup.html). ## Measurement Does Not Fix Anything I sell production tracking software, so this section is the disclaimer I would want to read. Knowing that machine 23 has been the bottleneck since 10 AM does not repair machine 23. A mechanic with a screwdriver does. Knowing an operator's output dropped after lunch does not raise it; a conversation might. Every fix on a sewing floor is physical and involves a person walking somewhere. No dashboard has ever sewn a garment. What changes is the delay. The same fact — machine 23 is the constraint — reaching the supervisor at 10:40 instead of at the end-of-day meeting is the entire difference, and it is worth several hours of output on that line alone. Multiply by the number of times a week something goes wrong on a floor, which is many. So the fair claim for any floor management system, ours included, is narrow: it shortens the distance between a problem starting and a supervisor knowing about it, and it removes the counting work that was keeping the supervisor off the floor. Everything after that is still people. A factory that installs tracking and expects output to rise on its own will be disappointed, in the same way a factory that buys a better pitch diagram and never re-runs it will be. The tool is only worth what someone does with it in the hour after it speaks up. ## Frequently Asked Questions ### What does sewing floor management actually mean? It means keeping a set of sewing lines producing at their planned rate by removing whatever is stopping them. In practice that is seven control areas: line balancing, work-in-progress on the floor, hourly output against target, in-line quality checking, needle and sharp-tool policy, machine condition, and the physical state of the aisles. The management part is not the list itself, which every factory knows. It is knowing, hour by hour, which of the seven is currently slipping. ### How does a supervisor track hourly output without a clipboard? Either someone walks the line and counts, or the count is a by-product of the work itself. Manual hourly counting works on one line and stops working across a floor of hundreds of machines, because the counting takes longer than the hour it is measuring. The alternative is to make the handover of each bundle a recorded event: a scan at the point where an operator picks up or finishes work, carrying lot, article, colour, size, component and quantity, so the hourly figure assembles itself. On our floor the scan records operator, machine, operation and timestamp, and nobody writes a tally. ### What is a bottleneck on a sewing line and how do you spot one early? A bottleneck is the operation whose hourly capacity is below the line's target, so it sets the output of the whole line. The visible symptom is a growing pile of work in front of one machine and idle hands after it, but by the time the pile is obvious you have already lost the hours it took to build. The earlier signal is the rate of change: work arriving at that operation faster than it leaves, for two or three hours running. That is a comparison of input and output counts per operation, which is why bottleneck detection and output tracking are really the same problem. ### Why is WIP sitting on the sewing floor a problem if the work still gets done? Because it is finished fabric, cut and partly sewn, that has been paid for and cannot be invoiced until it leaves as a garment. It also hides problems: a defect made on Monday is discovered on Thursday when the bundle finally reaches checking, by which time the same fault has been repeated across three more days of output. High WIP lengthens the gap between making a mistake and learning about it, which is the real cost, on top of the cash. ### Does production tracking software fix sewing floor problems? No, and any vendor who says otherwise is selling you something. Knowing that machine 23 has been the bottleneck since 10 AM does not repair machine 23; a mechanic does. What changes is the delay. The same fact reaching the supervisor within the hour, instead of at end of day or in the next morning's meeting, is worth several hours of production on that line. Measurement buys you time to act. It does not act. --- **Hourly Output Without the Clipboard** Scan ERP turns each bundle handover into a recorded event — operator, machine, operation, timestamp — so hourly output, WIP age and bottleneck position assemble themselves. Stalled work is flagged past six hours, stuck operations are swept every two minutes, and bottleneck summaries go out at 11:00 and 15:00. Built and run in a 500+ machine CMT factory across 50,000,000+ tracked piece-operations. From $5 per machine or $0.001 per piece, whichever bills lower, starting at $39/month. [Request a Free Demo](https://scanerp.pro/#contact) --- If you take one thing from this: next time you read a sewing floor management checklist, go through it and write beside each item how you would find out today, before lunch, that it had stopped happening. The items where you cannot answer that are the ones actually costing you money, and they are usually not the ones being discussed in the morning meeting. *Santosh Rijal is the founder of [Scan ERP](https://scanerp.pro/), a garment manufacturing ERP system designed for factory floor operations. He works directly with sewing lines, cutting rooms, and production supervisors across Nepal's garment manufacturing sector.* ## Related guides from the field - [WIP Tracking in a Garment Factory](/blog/wip-tracking-garment-factory.html) — Why work sitting on the floor is money, and how to see its age. - [Sewing Line Efficiency Calculation](/blog/sewing-line-efficiency-calculation-garment-factory.html) — The arithmetic behind the hourly target you are measuring against. - [Garment Factory Layout and Sewing Line Setup](/blog/garment-factory-layout-sewing-line-setup.html) — Aisle space, trolley routes, and the layout decisions that trap WIP. - [What an Industrial Engineer Does on a Garment Floor](/blog/industrial-engineer-garment-factory.html) — Who owns line balancing, capacity studies, and the method work.