Textile, Apparel and Footwear Manufacturing Case Studies
Industrial5 Case StudiesUpdated Sep 2026
Practitioners who work in fiber-to-garment and fiber-to-footwear production spend more time with machines, tolerances, and maintenance logs than with fashion calendars. The five cases below come from plants of different sizes, yet they share a pattern. A small mechanical or process detail drifts outside its design window and drags throughput, quality, safety, schedule, or resource use along with it. Each case uses the same frame of constraints, operating data, tolerances, and upkeep records, so readers can see how evidence was gathered and where judgment entered. All organizations are fictional, and every figure belongs to them.
Case Study 1
Stenter chain bushings wearing out ahead of schedule
Calder Mill Finishing, a family-owned finishing house with two stenter lines, about 160 employees, and a three-person reliability group.
The situation
Calder Mill runs two stenter frames that heat-set woven cotton and polyester blends for a regional shirting customer. Each frame carries two endless chains fitted with clips, and the clips grip the selvedge as the fabric passes through oven zones held near the upper end of the process window. Over the past several seasons the bushings in the chain links began to wear unevenly. The maintenance group had been replacing them on a fixed interval, but the interval had to be cut by almost half after the first failures. Operators reported edge curl and a slight skew that the finishing inspection caught only after a lot had already been dyed and cut.
The challenge
The plant needed to know whether the bushings were failing because of a material choice, a lubrication habit, or a tolerance stack that loaded one side of the chain. The supplier's drawing called for a bronze bushing with a specified clearance, but that clearance had never been checked against the chain's actual thermal growth. Nobody could say whether the failure was steady or sudden, and the records did not show lubricant quantities per shift.
Analysis
The first step was to put the design envelope beside the operating envelope. The bushing was rated for a given temperature, but the frame ran near the top of that range for long shifts. Oven readings drifted up over each shift, a pattern the chain had never been checked against. The drawing clearance was measured on a cold chain. Once the thermal growth of the link pins was added, the running clearance on the loaded side closed up. Maintenance records showed replacements clustering after high-speed runs, and oil top-ups were logged per shift rather than per hour.
Teams grounded in Textile, Apparel and Footwear Manufacturing usually begin by separating what the part was designed to do from what the line actually asks of it. Here the candidate failure modes were lubricant starvation, galling between bronze and a hardened pin, and thermal misfit. Each left its own wear signature, so the pulled bushings were sorted by pattern first. Three corrective options were on the table: a polymer-lined bushing rated for the higher zone temperature, a wider running clearance, or automatic oil dosing with a logged feed rate. The team preferred the combination, since the polymer option alone would not fix the starvation pattern.
Actions taken
- Log oil feed rate and oven zone temperature at each bushing change, so wear can be tied to specific operating windows rather than calendar intervals.
- Measure running clearance on a warm chain at operating temperature, and set the tolerance band from that measurement instead of the cold drawing value.
- Trial the polymer-lined bushing on one frame for a full production cycle, with the other frame kept on bronze as the control, and compare wear at matched intervals.
- Write a shift checklist for the oil dosing check and add a weekly review of wear photos, with the reliability lead signing off on any change to the interval.
Results
The polymer trial ran on the first frame through a full production cycle. Bushing replacements on that frame moved to a longer interval, and edge skew complaints stopped appearing in the finishing inspection log. The control frame stayed on bronze with logged oil feed, and its wear pattern shifted from scoring to even wear, which pointed to starvation rather than galling. The team did not claim the polymer as a cure. The clearance change alone had accounted for part of the improvement, and the two effects had not been separated cleanly.
Questions for discussion
- Which operating records would you need before deciding whether a material change or a lubrication change should come first?
- How would you separate the effect of a clearance change from the effect of a new bushing material on the same frame?
- What would make you distrust a wear measurement taken on a cold chain?
Case Study 2
Seam puckering throttles a jeans sewing line
Ridgeline Denim Works, a workwear sewing plant with four cut-and-sew lines, a six-person quality lab, and a shared maintenance crew of five.
The situation
Ridgeline Denim Works sews heavy cotton workwear on four cut-and-sew lines, with the rear yoke and side seams as the most scrutinized points. Over several weeks, output on the lines fell by about a fifth, and the rework bin at the end of the line filled faster than the repair station could clear it. The quality lab traced most of the returns to puckering along the yoke seam, a defect that looked minor on the table but added a re-sew cycle at the attach station. Supervisors first blamed the newest operators, then the fabric, and the argument went around the floor for a while without producing a fix.
The challenge
The plant had to decide which variable deserved the first attention, because each candidate pointed to a different fix. Needle and thread pairing, fabric lot shrinkage, feed dog wear, shift-to-shift tension settings, and individual method all had some support in the data. The difficulty was that no one had recorded them together. Acceptance criteria also had to be written before any change, so the plant would know whether a fix had worked.
Analysis
The line was mapped from cut panels to finished garment: prep, yoke attach, side seam, bartack, inspection, and repair. Yoke attach carried the highest defect rate and the lowest rated throughput, so it set the pace for the whole line. Operating data from the machine controllers showed that stoppages clustered in the first hour after a needle change and after a fabric lot switch. Puckering was judged against a visual reference panel, and the two shifts read that panel differently. Maintenance records showed presser foot and feed dog replacements logged only after failures, never on a scheduled basis.
Each variation source was scored by how much it could plausibly move the yoke station's cycle time and defect count. Needle and thread mismatch scored highest because the first hour of runs after a change was consistently worse, and the mismatch explained a large share of the early stoppages. Fabric lot shrinkage mattered because the same panel size came off different lots with different bow, which fed uneven tension into the seam. Feed dog wear and tension settings were real but smaller contributors. Individual method was the last thing considered, since treating it first would have punished operators for what was really a setup problem.
Actions taken
- Pair each needle size and thread type to a fixed setup card, and require a logged first-hour check before the station is released to production.
- Accept a fabric lot only after a panel shrinkage check and a bow measurement, with a pass threshold written into the receiving form before the first lot arrives.
- Replace feed dog and presser foot parts on a scheduled interval set from teeth wear measurements, and keep a record of each change against the lot running at the time.
- Set acceptance criteria: yoke station output recovers to at least 95 percent of the baseline rate for three consecutive weeks, and puckering rework falls to no more than 2 percent of panels inspected, measured on the same reference panel.
Results
Needle pairing and lot checks went in first, and the first-hour stoppage pattern shrank within the initial weeks. Yoke station output came back to about 96 percent of baseline by the end of the fifth week, and rework fell under the target on the lines that had adopted the setup card. Two lines lagged because their feed dogs were overdue for replacement, which the scheduled interval had not yet caught. Operators did not see a method change, which is the result the plant had wanted, since most of the loss had been a setup issue rather than a skills gap.
Questions for discussion
- Which records would you need to separate a fabric problem from a needle problem when both appear in the same first hour?
- How would you set a puckering acceptance threshold that two shifts apply the same way?
- What evidence would justify moving a maintenance interval from a fixed schedule to a wear-based one?
Case Study 3
Adhesive room review finds missing ventilation evidence
Kestrel Sole Works, a footwear assembler with two lasting lines and a proposed adhesive room for a new performance sole.
The situation
Kestrel Sole Works currently applies its cement at open benches on the main assembly floor, where lasting and trimming share the same air. The engineering team proposed a closed adhesive room for a new performance sole that uses a stronger solvent-based cement, with spray booths, a drying tunnel, and a single exhaust fan. Dani, the process engineer, submitted the project with a layout, a ventilation drawing, and the new cement's data. Rafael, the safety officer, reviewed the package before any purchase order could move. His first concern was that the file held a design airflow number but no measured capture at the bench where workers actually stand.
The challenge
The review had to decide whether the package was ready to proceed to purchase, and what evidence was still missing. Hazards included flammable vapor accumulation, electrical ignition sources in the room, worker exposure during spraying, and fire protection coverage for the drying tunnel. The package did not include a capture velocity test, a current safety data sheet for the new cement, an area classification, or a fire protection certificate. Each gap blocked a different approval gate.
Analysis
The first analysis tied the hazard to the design basis. Vapor generation depended on how much cement operators applied per shift, and the file estimated that from a single trial rather than from bench logs. Exhaust was sized for a fixed rate, but capture at each bench depended on hood position and spray pattern, and neither had been measured. Maintenance records for the existing floor fans showed belt changes but no motor rating, so those fans could not be reused in the room. The margin below the lower explosive limit at the breathing zone was never stated.
Documentation was the second layer. The local fire code's chapter on flammable liquids governs storage and handling quantities, and the room's classification and detection needed to follow from those quantities, not from the room's floor area. The plant's written chemical hygiene program required a current safety data sheet and an exposure assessment before a new substance reached a workstation, and neither was on file for the new cement. Each gate needed a named sign-off. The purchase gate could be cleared on conditions, since equipment could be ordered with a rated motor, but the operating release could not be granted until the capture test and classification were in hand.
- 1Hazard list and scopeVapor sources, ignition sources, and exposure tasks identified.
- 2Capture test evidenceMeasured bench capture must meet the design value first.
- 3Electrical classificationFan, lights, and sensors matched to the room's rating.
- 4Fire protection sign-offDetector and sprinkler coverage certified for drying tunnel.
- 5Operating releaseNamed sign-offs on the procedure before any production.
Actions taken
- Run a capture velocity test at every bench position with the new cement in use, and reshape the hood layout where capture falls short of the design value.
- Obtain the current safety data sheet for the new cement and compare its flammability data and exposure limits against the room's design basis before any purchase.
- Have a qualified electrical reviewer classify the room and confirm that the fan, lights, and sensors carry ratings that match that classification.
- Hold operating release until the fire protection contractor certifies detector and sprinkler coverage for the drying tunnel and the safety officer signs the revised procedure.
Results
The review cleared the purchase gate on conditions and held the operating release. The capture test found two of six bench positions below the design value, and the layout moved both hoods before the room was framed. The safety data sheet for the new cement listed a lower flash point than the earlier estimate, which pushed the area classification into a more demanding category. The fan was respecified with a rated motor. The project moved back by several weeks, and the team accepted that delay as the cost of a room that could show its evidence.
Questions for discussion
- Which evidence would you ask for first if a vendor offered a higher-capacity fan without a measured capture result?
- How would you decide whether a bench-level capture failure justifies redesigning the whole room?
- Where in your own approval process would a conditional purchase gate create the most risk?
Case Study 4
Knitting cell cutover slips after drive delay
Thornbury Knit Mills, a sweater and technical knit producer with three flat-knitting rooms and a growing automation program.
The situation
Thornbury Knit Mills had scheduled cutover of a new automated flat-knit cell to coincide with the start of a seasonal order from a retail customer. The customer's contract set a fixed delivery window, and the production plan had about four weeks of float built in before that window closed. The cell's carriage drives were ordered in the first week of the project, with a supplier quote that stated a lead time the procurement lead, Teodora, had accepted without checking against the supplier's current backlog. Electrical work was scheduled in parallel, with the rough-in done by a contractor who shared a crew with the building's other trades.
The challenge
By the middle of the project the drives were still in the supplier's queue, and the electrical rough-in had hit an existing cable tray that the drawings had shown in the wrong position. Four recovery options were on the table, and each consumed float or money in a different way. The milestone could not move without the customer's agreement, so the plant needed numbers it could put in front of the account manager.
Analysis
The first analysis separated the delay into causes. The drive choice rested on torque and speed, but the carriage's encoder alignment tolerance was tight, and the older rooms had never run that combination. Commissioning logs from the first two carriages showed encoder readings drifting when the cable run was long and loose. The tray clash came from comparing the as-built survey with the electrical drawing too late, after the rough-in crew was booked. Maintenance records showed spare drives on the older carriages, but none matched the new frame size, so no internal substitute existed.
Four recovery plans were compared on float used, cash cost, and customer exposure. Expediting the drives with premium freight cost $14,000 and recovered the most time, but only if the supplier's new date held. Splitting the first run to six carriages cost nothing directly but cut weekly output for that run. Overtime for the electricians cost $11,000 and cleared the rough-in, yet it did nothing for the drives. Asking for a later window avoided cost, but the contract's late-delivery clause carried $26,000 of exposure. The team chose expediting with the split.
- Start of projectDrive order placed on the supplier's quoted lead time.
- Week twoSupplier backlog pushes drives past the planned date.
- Week fiveRough-in hits a mispositioned tray the drawings missed.
- Week sixPartial first run and expedite plan approved by account.
- Week nineCell cutover completes with most of the float used.
Actions taken
- Get the drive supplier to confirm a firm ship date in writing, and release the premium freight only when that date is on the carrier's tracking record.
- Walk the cable tray route with the electrical lead and the as-built survey together, then reroute the tray along the north wall before the crew is rebooked.
- Commission the first two carriages with a shorter, tied cable run, and log encoder readings at each step to confirm whether the drift came from cable length.
- Tell the customer the first run will be partial, with a written weekly quantity and a revised full delivery date, and get the account manager's sign-off before production starts.
Results
The split and expedite plan went ahead. Drives arrived in the seventh week and the cell cut over in the ninth, which used most of the float but kept the first delivery inside the customer's window. The first run shipped a partial quantity each week, and the customer accepted the revised schedule in writing. Encoder drift disappeared once the cable run was shortened, which confirmed the cause. Overtime was not used, and total extra spend was $14,000 in premium freight.
Questions for discussion
- What would you check first when a supplier's lead time has not been verified against its current backlog?
- How would you weigh a partial delivery against a late full delivery when the customer contract is fixed?
- Which records would show whether a coordination problem could have been caught before the crew was booked?
Case Study 5
Heat recovery trial on a knit drying range
Marrow Creek Dyeing Co., a piece-dye house for cotton and cotton-blend knits, with one drying range and a wastewater treatment unit.
The situation
Marrow Creek runs one gas-fired drying range that dries dyed knit fabric after washdown and softening. The range had been run at a single profile for every shade family, and the gas bill was the largest single utility cost in the dye house. The operations manager, Priyanka, wanted to cut energy per kilogram of fabric dried and to reduce the water used in rinse cycles upstream, without changing the shade results or hand feel that buyers had come to expect. The first proposal was a heat exchanger on the exhaust stack, which would pre-heat make-up air. A second option was simply retuning the burners and reducing the profile for lighter shades.
The challenge
The decision was not simply which option saved more energy. Heat recovery could shift the drying temperature and moisture load, and the shade standard allowed only a narrow color difference per batch. Lint in the exhaust could also foul the exchanger faster than the maintenance plan covered. The team needed a comparison built on lifecycle energy, water, and rejects, with a clear quality limit that no option could break.
Analysis
The team began with operating data rather than the equipment catalog. Exhaust left the range hot and carried moisture that varied with shade family. Burner logs gave energy per kilogram for each shade family, and the deep shades used the most gas. The lifecycle comparison weighed the exchanger's manufacturing energy, replacement parts, and fouling losses against gas saved over its service life. Maintenance records showed weekly lint clearing already on the exhaust duct, so an exchanger would add a task rather than replace one.
Quality tolerance set the hard limit. The shade standard allowed a fixed color difference per batch, and past rejects had traced drift to drying temperatures that moved outside a narrow band. A recovery loop could move that band, so any option had to hold the drying profile within it. Hand feel and shrinkage mattered too, since fabric dried too fast or unevenly could change its stretch and dimensional stability. The design constraint was therefore not energy alone. Options were judged on whether they kept the profile inside the tolerances that shade and hand standards already required, and no option that broke that profile was accepted.
Actions taken
- Install temperature and humidity probes on the exhaust and make-up air, and record energy per kilogram for each shade family over the trial period.
- Run paired shade trials, one batch dried on recovered air and one on standard air, and measure color difference against the agreed batch limit.
- Add a pressure drop check on the lint filter to the weekly plan, with a threshold that triggers cleaning before the exchanger starts losing output.
- Approve capital only after the lifecycle comparison shows net savings once exchanger manufacturing energy, replacement parts, and fouling losses are counted.
Results
Recovered air worked well for lighter shades. Energy per kilogram on those runs dropped by about a quarter compared with the standard profile, and hand feel matched the reference. Deep navy batches drifted outside the color limit in the paired trial, so the team kept standard air for dark shades and logged the split by batch. The weekly lint task took a short time once the pressure check was in place. The project moved forward as a partial recovery system, not the full range.
Questions for discussion
- Which measurements would you collect before deciding whether a heat recovery system should cover every shade family?
- How would you decide whether a fouling penalty is small enough to ignore in a lifecycle comparison?
- What quality evidence would convince a buyer that a changed drying process has not altered their product?
Learning Resources
Industrial CASE STUDIES
China, India, United States, Indonesia, Pakistan, Nigeria, Brazil, Bangladesh, Russia, Japan, Mexico, Philippines, Vietnam, Egypt, Turkey, Germany, United Kingdom, France, Italy, Canada, South Korea, Australia, Spain.