Progressive Bundle System (PBS) Layout & Work Flow

The progressive bundle system is the oldest work flow in garment manufacturing, and it is still the work flow most factories run. Cut fabric arrives at the sewing floor in bundles, each bundle carries a ticket, and each bundle moves from operation to operation until it reaches finishing as a finished garment bundle. The system is simple, cheap to start, and forgiving of mixed styles. It is also easy to run badly, because its failure mode is silent: the floor looks busy, every operator has work, and yet shipments leave late and defects are found days after they were sewn.

This lesson explains how PBS works, how to choose a bundle size, how to cap work in progress, and how to measure whether the line is actually flowing. The search phrase that brings most readers here is progressive bundle system in garment manufacturing, and the answer they need is not a definition but a discipline: bundle size by arithmetic, work in progress by cap, and traceability by ticket.

A bundle is a fixed number of cut pieces of one style, one size, and one shade lot that travel together. The bundle is the unit of transport, the unit of counting, and the unit of control. Operators do not receive loose pieces; they receive bundles. Supervisors do not count loose pieces at shift end; they count bundles. Payment, progress, and traceability all follow the bundle, because counting 400 bundles is reliable while counting 10,000 loose pieces is not.

The bundle ticket is what makes the bundle a control unit rather than a pile of cloth. A good ticket shows style number, size, shade lot, quantity, cut number, and the operation sequence with a box to tick at each operation. When the ticket travels on top of the bundle, in the same place every time, any operator can answer five questions at a glance: what is this, how many, which shade, what is next, and what is already done. When the ticket is missing, buried, or wrong, the bundle is just cloth, and the factory has lost control of it.

Bundle size is a trade-off, and the trade-off can be calculated. Start from the order and the line. In the worked example used through this lesson, a knit factory receives an order for 10,000 polo shirts and sets the bundle size at 25 pieces. Dividing 10,000 by 25 gives 400 bundles. The sewing line target is 1,000 pieces per day, which is 40 bundles per day, so dividing 400 by 40 gives a schedule of 10 working days. Those three divisions are the whole planning exercise, and they are worth doing before cutting starts, because they turn a vague order into a countable flow.

Small bundles flow faster and reveal defects sooner. A bundle of 10 pieces reaches the next operation quickly, and a defect found in it costs at most 10 pieces of rework. But small bundles cost handling: more tickets, more counting, more trips between operations, and more chances for a bundle to be misplaced. Large bundles cut handling cost, but they delay feedback: a bundle of 100 pieces can carry a defect through five operations before anyone notices, and then the rework is 100 pieces, not 10.

The practical starting range for most woven and knit lines is 20 to 25 pieces. From there, adjust by measurement, not by opinion. If bundle throughput time is long and rework lots are large, go smaller. If operators spend more time fetching bundles than sewing, go larger. Change one variable at a time and measure for two weeks, because bundle size interacts with work in progress and line balance, and changing everything at once teaches nothing.

Work in progress is the number of bundles on the floor that have started but not finished. It is the single most useful health number on a PBS line, because it exposes bottlenecks that output figures hide. A line can hit its daily piece target while drowning in work in progress, and that line will ship late the week a style changes or a shade problem appears.

The discipline is a cap: a maximum number of bundles allowed per operator on the floor. In the worked example, the line has 25 operators and the cap is 2 bundles per operator, giving a floor cap of 50 bundles. Before the cap, the floor held 120 bundles; after the cap, it holds 50. The reduction is 70 bundles, which is 58.3 percent of the old level. The mechanism matters more than the number: when the cap is reached, the line stops releasing new bundles and fixes the slow operation instead of feeding it more work. Feeding a bottleneck only deepens the pile and lengthens every bundle's wait.

Counting is what enforces the cap. At the end of every shift, the supervisor counts floor bundles by operation and compares the count with the cap. Excess bundles go back to the cutting buffer, not to the next operator's table. This feels wasteful the first week, because operators see empty tables and ask for work. Within two weeks, the effect reverses: operators stop hunting through piles for the next bundle, throughput time falls, and the line's output rises on the same headcount.

Throughput time is the time from a bundle entering sewing to the same bundle leaving finishing. In the worked example, a bundle passes 25 stations, and before discipline each station holds it in queue for an average of 30 minutes. Multiplying 25 by 30 gives 750 minutes, or 12.5 hours, per bundle. After discipline, the average queue falls to 12 minutes, and 25 multiplied by 12 gives 300 minutes, or 5 hours. The saving is 7.5 hours per bundle, and it comes from queue reduction alone, with no change to sewing speed.

Handling is the second proof. Before discipline, each bundle is touched 12 times across the order: fetching, counting, moving, recounting, and stacking. For 400 bundles that is 4,800 touches. After discipline, touches fall to 8 per bundle, giving 3,200 touches, a saving of 1,600 touches. At 0.5 minutes per touch, the saving is 800 minutes, or 13.3 hours, across the order. Neither number is dramatic on its own; together they explain why a line that sews at the same speed can ship days earlier once its flow is disciplined.

PBS layout follows the unidirectional flow principle from Module 1: material moves forward from cutting to sewing to finishing without backtracking. Within the sewing floor, each operator needs an input side and an output side, clearly marked, so bundles never pile on the machine table itself. The input side holds the next bundle; the output side holds the finished bundle for collection. A bundle collector, sometimes called an input man, moves finished bundles forward on a fixed round, rather than letting operators leave machines to carry bundles themselves.

Buffer space belongs between departments, not between every operator. A small buffer before a known slow operation, such as a buttonholing or embroidery station, absorbs variation without flooding the floor. A buffer after every operation is just work in progress with a nicer name. Mark buffer zones on the floor with paint or tape, label the maximum bundle count for each zone, and treat an overflowing buffer zone as a signal to rebalance, not as normal.

A ticket system fails in predictable ways, and each failure has a simple countermeasure. Tickets fall off: staple or pin them, and put them in a transparent pouch on top of the bundle. Tickets get wet or torn: print on card, not thin paper, and keep bundles off the floor. Operators tick the wrong box: print the operation sequence in the order the bundle actually travels, and number the boxes. Shade lots get mixed: print the shade lot in large type and colour-code the ticket edge by shade family.

The rule that matters most is the splitting rule. A bundle may only be split when a new ticket is created for each part, showing the split quantity and the original cut number. A split without a new ticket is how a size run goes missing: the original ticket says 25 pieces, the line holds 15, and nobody can say where the other 10 went. Enforce the rule without exceptions, including rush orders, because rush orders are exactly when traceability matters most.

PBS suits bulk production of stable styles: long runs, standard operations, and lines that change style weekly rather than daily. It suits short runs and samples poorly, because bundle changeover, ticket work, and counting overhead eat the small quantity. It suits single-piece flow products poorly too, because the bundle itself is the delay: a garment cannot finish until its whole bundle finishes each operation.

Factories therefore run PBS for bulk and other systems for the rest. Samples and short runs often go to modular teams, and high-volume stable styles sometimes graduate to unit production systems. Knowing when PBS is wrong is part of knowing PBS, and the signal is consistent: if bundle throughput time exceeds one working day for a style with fewer than 20 operations, the bundle is the bottleneck, and a smaller bundle, a modular team, or a single-piece line deserves a trial.

Start with one line, not the whole floor. Measure its current bundle count, floor work in progress, and throughput time for one week. Choose a bundle size from the order arithmetic, print the new ticket, paint the buffer zones, and set the per-operator cap. Count bundles at every shift end for two weeks, and publish the three numbers on the line board: bundles completed against target, floor bundles against cap, and average throughput time against the 5 hour standard used in this lesson. When all three hold for two consecutive weeks, roll the system to the next line, and keep the old line's numbers as its baseline for the next improvement.

Progressive bundle system discipline is unglamorous work: counting, capping, ticketing, and measuring. It is also the cheapest productivity improvement available to a garment factory, because it needs no new machines, only the decision to stop feeding bottlenecks and start counting bundles.