The importance of inspection and testing for producing quality finished garments is well established. Though 100% inspection systems are prevalent in apparel factories, it is also a known fact that 100% inspection is not always 100% perfect, and which also adds to additional manpower costs. On the other hand, online quality control systems based on statistical sampling plans can ensure conformance to quality requirements. AO Bangladesh highlights online quality control systems for an apparel manufacturing factory.
Online inspection is inspection of parts and partially assembled products as they are progressively made into complete products. In apparel manufacturing, this means inspection at various points in the entire manufacturing process from spreading fabric to pressing/finishing. The idea behind in-process inspection is to inspect or check the quality of workmanship and component parts as close to manufacturing processes as possible and thereby identify the source of abnormality and take corrective and preventive actions as early as possible in the manufacturing process.
Designed to uncover deficiencies in workmanship as well as equipment malfunction, online inspection can be performed by either quality control inspectors or individual operators themselves after they perform their respective operation(s). This will result in minimising the need for later repairs and re-work.
Similar to online inspection, online testing involves tests for checking the seam strength, washing fastness of the sewing threads and labels, tests for the strength of the zippers and buttons. Instrumental in ensuring conformance to the buyer’s stated quality standard, online tests strive to eliminate deviations from the required quality norms and their sources as early as possible.
Online Inspection in Cutting Room
Spreading and cutting can arguably be termed as operations for which clear quality inspection guidelines and standards are missing. Each spread must be inspected and the internal procedures should be such that unless a spread is inspected and passed by the online inspector, cutting should not commence. Similar parameters should be kept in mind for operations of fusing and printing, which are either a part of or subsequent to the cutting room. A sample checklist for the cutting room processes has been given in Exhibit 1.
Online Inspection in Sewing Room
In-process/online inspection in sewing involves the inspection of work from each operator, with a quality standard established to limit the amount of bad work permitted and a provision for operators to re-inspect and repair entire bundles should this limit be exceeded. The decision on where to place inspection stations is influenced by various factors, such as the importance of operations and controlling troublesome or key operations. First, a complete manufacturing process chart (Picture 1) should be made, clearly identifying the production or manufacturing steps for each type of garment made.
Thereafter, inspection points or stations should be carefully selected so that the operations to be checked are neither covered by later operations, nor necessitate ripping good work to repair a defect. Inspection stations should provide a uniform workload for each inspector and minimize the elapsed time between the completion of an operation and its inspection. Each operator should be told what standard of work is acceptable and what is not.
It is advisable to plan on having enough inspectors so that no operation ever goes longer than 4 hours without being inspected. Depending on the results of such inspection for at least 30 days the time interval of 4 hours can be either increased or decreased. The daily volume of garments produced should decide the numbers of inspectors and not the number of operators.
Typically, a small number of operators will cause the majority of the defects. Most inspections online can be performed rapidly without sacrificing accuracy and quite large units of inspection should be expected from inspectors each day. The workload of inspectors should be paced and also supervised. All the same, they must be allowed enough time to inspect in greater detail when necessary and to carefully explain quality problems to supervisors. The thoroughness of inspection is more important than the quantity inspected.
Table 1: Heiland’s Skip Bundle Sampling Plan | ||||
Sl. No. | Number of samples to be inspected from a bundle (n) | Number of defects allowed in n samples (c) | Skip interval (s) | Clearance interval (m) |
1 | 12 | 0 | 4 | 2 |
2 | 30 | 1 | 4 | 2 |
The selection of samples from a bundle must be at random. The number of samples selected for inspection should be based on statistical sampling plans. One such sampling plan given by Heiland, namely skip bundle sampling is shown in Table 1. For an example, we select a bundle at random at an inspection point and inspect 12 pieces (n) out of that bundle also at random. If no defective piece is found (c), accept that bundle and then inspect every fourth (s) bundle. If one or more defective pieces are found, reject the bundle and return it to the operator through his or her supervisor. The operator would then sort out the bundle and repair the defective pieces. These repaired pieces are inspected and if acceptable, the bundle moves on to the next operation. Once a bundle is rejected, every bundle from that operator is inspected from then on until two (m) consecutive bundles from that operator are found acceptable. From that point, revert to inspection of every fourth bundle.
For a unit production system, where an operator works on one garment at a time as it comes along a track or rail and each operator continues to perform his or her own special task, but individually instead of by the bundle, continuous sampling is called for. For example, a simple continuous sampling plan called CSP-1 works as mentioned in the following example. Assume an average outgoing quality level (AOQL) of 3.
1. In the beginning, inspect all the units or pieces consecutively as produced and continue until 50 pieces in succession are found acceptable.
2. When 50 pieces in succession are found acceptable, discontinue the inspection of all pieces and inspect only 1 out of 20 pieces produced.
3. If a piece is found to be defective, revert immediately to inspecting 50 pieces in succession and continue until again 50 pieces are found acceptable in succession. Then resume sampling inspection.
4. All defective pieces should be either repaired or replaced with good pieces.
Besides taking into consideration defects, ranging from the popular defects like needle damage, feed damage, skipped stitches, seam grin, pleated seam to insecure back stitching, companies should come up with their own quality and workmanship standards based on their product specialisation and make them a part of their quality manual. It is important that apparel retailers/brands develop their quality/workmanship standards keeping in mind the end consumer perspective than trying to invent as many defects as possible.
Online testing in the sewing room tests the garments for parameters like seam strength. A sample picked randomly from the lot as per the AQL is tested for seam strength through pull testing. The parameter to be observed in this case is the pulling strength. Zippers are also pull tested at this stage and observations are recorded for the pull strength at the point when either the fabric or the zipper was damaged.
End of Line Audit
Once the sewing is done, the garments are checked as per the specifications based on the dimensions of the garment at the critical positions and the placements of the labels. For example, if the lot size is 60 pieces, the respective sample size code letter will be E (refer to Table 2). Corresponding to this, as per Table 3, a sample size of 13 pieces is to drawn at random. If the AQL is set at 2.5, a maximum of 1 defect can be accepted and anything more than that will lead to the rejection of the entire lot of 60 pieces.
The AQL level to be used is usually recommended by the buyer. However, if there is no recommendation, then the manufacturer must set an AQL level lower than the final audit AQL. This will ensure that the final audits do not fail because of the high rate of defects that could have been accepted during in-process auditing with a higher AQL.
Table 2: Sample Size Code Letters | |
Lot or Batch Size | Sample Size Code Letter |
2 to 8 | A |
9 to 15 | B |
16 to 25 | C |
26 to 50 | D |
51 to 90 | E |
91 to 150 | F |
151 to 280 | G |
281 to 500 | H |
501 to 1200 | J |
1201 to 3200 | K |
3201 to 10000 | L |
10001 to 35000 | M |
Online Inspection & Testing in Pressing & Finishing
The basic objective of finishing must be quality and appearance. Appearance is the basis of most consumers’ judgment on whether or not to purchase a garment. The quality of a pressing operation can be measured by inspecting and evaluating aspects like burnt or scorched garments, gloss and/or change in colour (original shade), flattened nap or surface, fabric of finished garment not smooth, wrinkle-free, and showing its proper appearance, garments not thoroughly dried and shrinkage due to heat and moisture.
The pressing and finishing room also includes due diligence procedures for stain/fading (buttons, trims, zippers, wash care labels) for wash programs and bond/stitch strength (buttons, rivets, fasteners) are checked as per the buyer’s recommended standards (frequency and sample size).
Table 3: Sampling Plans | |||||||||
Sample Size Code Letter | Sample Size | Acceptable Quality Level | |||||||
2.5 | 4.0 | 6.5 | 10.0 | ||||||
Ac | Re | Ac | Re | Ac | Re | Ac | Re | ||
A | 2 | 0 | 1 | 0 | 1 | 0 | 1 | 1 | 2 |
B | 3 | 0 | 1 | 0 | 1 | 0 | 1 | 1 | 2 |
C | 5 | 0 | 1 | 0 | 1 | 0 | 1 | 1 | 2 |
D | 8 | 0 | 1 | 1 | 2 | 1 | 2 | 2 | 3 |
E | 13 | 1 | 2 | 1 | 2 | 2 | 3 | 3 | 4 |
F | 20 | 1 | 2 | 2 | 3 | 3 | 4 | 5 | 6 |
G | 32 | 2 | 3 | 3 | 4 | 5 | 6 | 7 | 8 |
H | 50 | 3 | 4 | 5 | 6 | 7 | 8 | 10 | 11 |
J | 80 | 5 | 6 | 7 | 8 | 10 | 11 | 14 | 15 |
K | 125 | 7 | 8 | 10 | 11 | 14 | 15 | 21 | 22 |
L | 200 | 10 | 11 | 14 | 15 | 21 | 22 | 21 | 22 |
M | 315 | 14 | 15 | 21 | 22 | 21 | 22 | 21 | 22 |
Final Inspection
Final inspection consists of inspecting finished garments from the consumers’ point of view; size measurement; form fitting (putting garments on the proper-size mannequins to see if they properly fit the labelled sizes); and live modelling (having appropriate size individuals actually try on garments) if necessary, again to see if the garments properly fit the labelled sizes. Form fitting and live modelling will contribute significantly toward assuring that the garments in question fit right, look right, and feel right for their intended sizes.
Final inspection may occur before or after garments are packed in poly bags and boxes. If it is done after garments are packed, then proper size and style markings on the package can also be checked. In any case, there should be a list of points to be checked in a garment, including a table of finished measurements.
Control of proper size is extremely important because no matter how well a garment is made, unless it fits properly, it will be useless from a consumer’s point of view. Control of garment dimensions will not guarantee that the garments will fit right, look right, and feel right. Fit is dependent on pattern shape, proportion, and dimension as well as proper grain. Poor cutting and inaccurate sewing can result in poor or no fit at all.