Digital tracking for work in progress
Operations

Digital tracking for work in progress

Financial impact

$6,720USD/ yr EBITDA

Operational impact

3supervisor-hrs / day back
high confidence

The math

Supervisor time recovered$6,720/ yr

1.5 hrs/day × 2 supervisors × 280 days × $8/hr

Capex

$5,000

Payback

9 months

Difficulty

3 / 5

Time to implement

~8 weeks

Supervisor time saved / day
hrs
Supervisors affected
Tablets
$
Barcode printer
$
Mounts + scanners
$
Floor Wi-Fi
$
Tracking software (yr 1)
$

Also uses working days / yr 280 days, supervisor wage $8/hr. Edit on the dashboard.

Description

Today the factory tracks half-finished orders by hand. After fabric is cut, the pieces for a batch of identical garments, say 10 or 20 shirts, are tied into a bundle and carried between sewing stations in a brown paper bag with a paper ticket taped to the outside. Nobody has a live view of where any order actually is. To answer a simple question like where is the Al-Futtaim order, a supervisor walks the floor and reads the tickets on dozens of look-alike bags, some of which operators have stashed under their tables. The video shows more than 50 of these bags in play at once. This is a build-to-order shop, one that only makes what customers have already ordered, so the thing you are really selling is hitting the promised ship date, and you cannot protect a date you cannot see.

Evidence

  • [00:01:13] 17 paper bundle bags clustered on the floor by the pressing area
  • [00:02:55] Bundle bags stashed under workstations, out of sight
  • [00:04:08] A supervisor walking the floor with a phone and a pen, hunting for an order
  • [00:00:01, 00:03:25] No computers or Wi-Fi on the floor, just a time clock at the door and one laptop upstairs

Deep dive

Expand for detail

Load-bearing, card breaks if wrong

Operators will scan every bundle, every time.

The live screen is only as true as the scans behind it. If people scan 90% of the time instead of 100%, the missing tenth shows up as bundles frozen at a station that everyone knows physically moved, and a supervisor now spends more time chasing these ghost bundles than they ever did with paper. Compliance has to be near total, which is why pay eventually gets tied to the scan, so a skipped scan is a skipped payment. What it looks like on the floor if this is wrong: the screen says a station is starved of work, but a walk down the aisle finds five bags stashed at the operator's feet.

Managers will open the screen and act on it, the same day.

The data is worthless if it sits unopened while decisions still get made by walking the floor. The whole point is to catch the bottleneck during the shift and rebalance the line while it still matters, not to read a report after the order already shipped late. If supervisors default back to the eye test, the tablets become wall ornaments and the paper bags quietly return. What it looks like if this is wrong: the owner asks where an order is, and the supervisor walks off to go look instead of glancing at the screen.

Late deliveries are a real, recurring cost today.

This is where the true value lives, not in the soft $7K. The case for spending $5K rests on the factory genuinely losing money to late custom orders now: penalty charges, discounts to keep a customer, or air-freighting finished garments to catch a ship date. If orders almost always go out on time already, the main prize is not there and the card leans on soft supervisor time. What it looks like if this is wrong: the dispatch log shows on-time delivery, and nobody can remember the last rushed air shipment.

Supportive, card weakens but survives

The floor Wi-Fi can actually reach every scan point.

The tablets sync over Wi-Fi, and the floor has none today, just an office router behind dense metal racking that blocks the signal, with the pressing and quality room walled off in the corner. If coverage is patchy, scans queue up and upload late, so the screen shows stale positions and people stop trusting it. What it looks like if this is wrong: tablets flashing offline warnings, and one walled-off station chronically missing from the screen.

The cheap tablets survive the floor for a few years.

To stay under $5K the plan uses ordinary Android tablets in tough cases, not $1,000 rugged units. A sewing floor has lint, oil, dropped scissors, and steam-iron heat. If a tablet cracks or dies every few months, replacements eat the payback. What it looks like if this is wrong: a spider-cracked screen at the cutting exit and a station that keeps dropping off the board.

The screen works without much reading.

The floor workforce is mostly from India, Pakistan, and Bangladesh, and dense English menus will slow people down or get ignored. The screen has to lean on big icons, color (green for go, red for a defect), and simple numbers, not paragraphs of English text. What it looks like if this is wrong: operators poking hesitantly at the tablet, or a queue building while one person figures out which button to press.

Operators skip the scan under deadline pressure

Risk

Scanning feels optional the moment a rush order is on. People go faster by skipping it, exactly when you most need to see where the order is, and the missing scans turn the live screen into fiction.

Early Warning

The screen shows bundles frozen at a station for hours that you know physically moved, and the number of operators typing bundle numbers in by hand keeps creeping up.

Mitigation

Tie pay to the scan once the data is trusted, so skipping a scan means skipping getting paid. The scanner becomes the cash register.

Stake

The whole $5K plus the on-time-delivery prize evaporate: you are paying for a screen that lies.

Operators game the scan to smooth their pay

Risk

Because scans drive piece-rate pay (the per-bundle pay each sewer earns), operators may scan bundles they have not finished, or hoard a stack and scan them all at once, to make their output look steady.

Early Warning

Scan timestamps bunched at the end of the shift, or one operator's hourly output that looks impossibly even across the whole day.

Mitigation

Spot-audit: physically match a few scanned bundles against work actually completed, and flag any operator whose scans all land in one burst.

Stake

Payroll paid for work not done, plus productivity data too corrupted to trust for balancing the line.

Managers keep walking the floor and never open the screen

Risk

Supervisors trust their eyes over a tablet and keep doing the eye test, walking 20 meters to check the line instead of glancing at the screen. If they do not use the data, they will not enforce scanning either, and the whole thing quietly dies.

Early Warning

The owner asks for an order's status and the supervisor walks off to go look, instead of pulling it up on the screen.

Mitigation

Management stops accepting verbal "I saw it on the floor" updates and asks for the screen in every status check, so using it becomes the only way to answer.

Stake

The entire investment: an unused screen changes no decisions and saves no time.

The paper bags and floor conditions kill the barcodes

Risk

Bundles ride in brown paper bags that get crushed, oil-stained, and stashed under tables, and the dense metal racking drops the Wi-Fi the tablets need. A paper label stapled to a soft bag is unreadable by the time it reaches pressing.

Early Warning

Operators typing bundle numbers by hand because the scanner will not read the label, and one walled-off station chronically missing from the screen.

Mitigation

Print on synthetic (plastic) label stock, clip it to a rigid tag on the bag handle, and add a mesh access point or a wired drop wherever the survey found a dead spot.

Stake

Manual typing takes many times longer than a scan, so operators quit using it, and the blind spots hide the very bottlenecks you bought the system to see.

A. What to actually buy

You do not need one tablet per machine or a brand-name apparel suite. You need a handful of scan points and the cheapest software tier that does the job. Everything below fits inside the $5K first-year budget.

Software (pick the small-factory tier)

  • Scan ERP advertises from about $5 a machine a month, or fractions of a cent per bundle. For this floor that is on the order of $100 a month. It has a phone app operators use on their own Android phones with no install, which can drop hardware cost toward zero.
  • Odoo (manufacturing app): roughly $1,800 a year, open-source and highly customizable, with strong Hindi and Arabic translation. A solid budget default.
  • A purpose-built garment tool (Stitch-MES, Res.Q, TrackIT, WiMetrix and similar) sits a step up, roughly $2,500 to $3,000 a year, and is designed around bundles and per-bundle pay out of the box.
  • Avoid the brand suites. WFX ($15K to $60K a year), Apparel Magic ($145 to $1,505 a month), and Coats FastReact ($20K and up) are built for fashion brands managing design and finance, not a job shop tracking bundles. They are the wrong tier and break the $5K budget on their own.

Hardware for a handful of scan points

  • Tablets: 4 to 6 ordinary Android tablets in tough cases at about $200 each, one at each scan point, not one per machine. Rugged units with a built-in scanner (Zebra ET51) are nicer but $900 to $1,300 each. Cheapest: staff scan on their own phones.
  • Label printer: one thermal printer (about $200 to $300) loaded with synthetic (plastic) label stock. A paper label on a soft bag does not survive the floor.
  • Scanners and mounts: Bluetooth handheld scanners plus a swing arm that holds each tablet at waist height, so a scan takes about 3 seconds with no bending or setting the bag down.
  • Wi-Fi: the floor has none today. Budget about $1K for mesh access points, and expect to run a wired cable to the walled-off pressing and quality corner, where the metal racking kills the signal.

B. Why a whiteboard comes first, and when digital finally pays

This is the honest heart of the card. The cheapest way to "know where every order is" is not a tablet, it is a large magnetic whiteboard: each live order is a card that moves across four columns (Cut, Sewing, Quality, Pack), updated by hand two or three times a shift. Off-the-shelf factory boards run about $200 to $1,500, and they capture much of the value with none of the scan-discipline risk.

Why start there? Paper and eyeball tracking does not truly collapse until somewhere between 50 and 100 sewing operators. This floor, with 41 sewing tables, sits just under that line: busy enough to feel the pain (more than 50 bundle bags in play at once), but not so big that one supervisor plus a board cannot hold it. A whiteboard people actually keep current beats a digital system they quietly bypass.

Go digital only when all three are true:

  1. A two-point barcode pilot (cutting exit and quality control) proves the scan rate stays above 95% for 60 straight days.
  2. Managers are opening the live screen and acting on it every day, not just walking the floor.
  3. Order count or complexity is rising, so the whiteboard is genuinely always out of date.

If pilot scanning stays under 90%, stop. The whiteboard is your answer, and forcing the tablets will just buy you a screen full of fiction.

C. How to make people actually scan (adoption runbook)

The system is a habit, not a gadget. Protect the habit:

  • Training: about 30 minutes hands-on per operator, with a supervisor signing off on the operator's first five clean scans.
  • Floor support: in the pilot's first week, IT or a supervisor stands at the scan point and physically hands over the scanner for every bundle. Second week, on call and watching.
  • Ergonomics: mount every tablet at waist height on a swing arm, so an operator holding a bag can scan in about 3 seconds without bending or setting the bag down. Get this wrong and adoption dies on day one.
  • Reward, do not punish: run a weekly scan-completion contest by zone (Cutting vs Sewing vs Finishing) with a visible tally and a small grocery voucher for the winning zone. Do not dock pay for missed scans in the first 30 days; instead reward people for flagging a torn label so the reprint process gets tested.
  • Sustain it: zip-tie a laminated three-step picture guide (photos, no text) beside every tablet, since the floor is mostly Hindi, Urdu, and Bengali speakers.

What we did not have in the corpus and would request from the client on day one. The aggregated gaps across all 15 cards become the engagement-letter ask sheet.

  1. 01

    Shipping and dispatch log (last 12 months)

    Owner

    Sales / Dispatch

    How many custom orders shipped late and what each one cost: penalty charges, discounts to keep the customer, or air-freighting finished garments to catch a date. This is the real prize the card rests on, far more than the $7K of supervisor time.

  2. 02

    Supervisor time-study (3 days)

    Owner

    Operations Manager

    Have the yellow-shirt supervisor log his hours for three days to show how much of the day is really spent walking the floor hunting for bundles. This is the check on the 1.5-hours-a-day saving.

  3. 03

    Floor Wi-Fi coverage map

    Owner

    IT / Facilities

    Whether the office router reaches the scan points, or whether roughly $1K of the budget must go to mesh access points and a wired cable to the walled-off pressing and quality corner.

  4. 04

    Lost or misplaced bundle log

    Owner

    Floor Supervisor

    Any record of bundles that went missing between cutting and sewing, how often, and the rework cost. It sizes a pain the whiteboard and the barcodes both fix.

  5. 05

    Order count and lines running

    Owner

    Operations Manager

    How many orders and sewing lines run at once, to confirm the floor is still under the size where paper tracking fully breaks (50 to 100 operators), and therefore whether the whiteboard suffices for now.

  6. 06

    Operator language and literacy profile

    Owner

    HR / Floor Supervisor

    The main languages spoken on the floor (Hindi, Urdu, Bengali, Arabic, English), to choose software with the right translation and to design an icon-based screen people can use without reading English.