AQL 2.5 Backpack Inspection: UK Buyer's Guide

AQL 2.5 Backpack Inspection: The Direct Answer

AQL 2.5 is an acceptance quality limit: the maximum percentage of defective units a buyer accepts in a lot, sampled and judged under ISO 2859-1. For a 500-piece backpack lot at general inspection level II, the code letter is H, the sample size is 50 pieces, and a lot passes with three or fewer major defects and five or fewer minor defects, failing at four majors or six minors. Critical defects - compliance failures, sharp points, a failed nickel-release test - have an accept number of zero regardless of the AQL. The number is not a promise of quality; it is a statistical filter. A lot with an underlying 2.5% major defect rate has roughly a 71% chance of passing the 3-of-50 plan, which means AQL 2.5 does not guarantee that 97.5% of your goods are good - it guarantees that most bad lots are caught. Buyers who need tighter control specify AQL 1.0, which on the same lot uses a sample of 50 with accept 1 and reject 2 for majors, and catches roughly 92% of 2.5% lots. What makes AQL work in practice is the defect dictionary written before inspection.

Without it, a 3 mm print misregistration is a judgement call at the factory gate at 17:00 on a Friday. With it, the inspector ticks a box. Write the dictionary at tech-pack stage, agree it with the factory, and reference it on the purchase order alongside the golden sample number and the tolerances.

How ISO 2859-1 Sampling Actually Works

ISO 2859-1 is a sampling system indexed by lot size and inspection level. The lot size sets a code letter; the code letter plus the AQL set the sample size and the accept/reject numbers. For 500 pieces at level II the letter is H and the sample is 50. For 200 pieces the letter is G and the sample is 32. For 1,000 pieces the letter is J and the sample is 80. The tables are published by the International Organization for Standardization and the inspector must not choose the sample size by eye.

Three inspection levels exist. Level I uses a smaller sample and is used when the cost of inspection is high relative to the risk; level II is the default; level III uses a larger sample and is specified when discrimination matters more than cost. For a first order, or any order below 300 pieces, specify level III, which on a 200-piece lot raises the sample from 32 to 50 and catches the warm-up defects that dominate short runs.

Switching rules matter on repeat orders. ISO 2859-1 defines normal, tightened and reduced inspection: two consecutive rejected lots move you to tightened inspection with a larger sample, and five consecutive accepted lots can move you to reduced with a smaller one. State in the contract that inspection starts at normal and moves to tightened after any rejection. That single clause is the cheapest quality lever available to a British buyer on a repeat programme. Abrasion and coating test methods referenced in a defect dictionary are published by the American Association of Textile Chemists and Colorists.

Sample Sizes and Accept Numbers for Backpack Lots

The table below gives the practical numbers a British buyer needs for the common lot sizes, at AQL 2.5 for majors and AQL 4.0 for minors, which is the usual pairing in soft goods. Note that the minor AQL is looser than the major AQL: a cosmetic issue is tolerated at four times the rate of a functional one, because a 3 mm print misregistration does not cause a return and a failing zipper does.

Read the 500-piece row and plan around it. Fifty pieces are pulled at random from the finished lot, examined against the golden sample and the defect dictionary, and judged: three majors or fewer passes, four fails. A failed lot goes to 100% re-inspection at GBP 1.20-3.60 per unit, or GBP 600-1,800 on a 500-piece run, charged to whoever the contract says pays. Write that clause into the purchase order before the inspection, not after the rejection.

Critical, Major and Minor: Writing the Defect Dictionary

Critical defects are safety and compliance failures, with an accept number of zero. For a backpack they are: a sharp point or edge that fails a tactile check; a small-part detachment that fails a 90 N pull test on a component fitting inside a 57 x 32 mm cylinder; a nickel-release result above 0.5 micrograms per square centimetre per week; an azo dye result above 30 mg/kg; a phthalate result above 0.1% by weight; a missing or false UKCA declaration; and a cord length above the permitted limit on a children's product.

Major defects are functional failures that a British customer would return. Set the thresholds in numbers: a zipper that fails before 200 cycles or needs more than 1.4 kg of pull force; a strap attachment below 180 kg straight pull; a dimensional deviation above plus or minus 5 mm; a colour deviation above Delta E 2.5 under D65; a seam leakage showing a damp ring within 30 minutes; a sleeve that does not clear a 355 x 250 mm device; and a weight deviation above plus or minus 4%.

Minor defects are cosmetic and tolerated to a limit. Set them too: a print misregistration above 1.5 mm up to 3 mm; a loose thread of 8-20 mm; a dimensional deviation of 2-5 mm; a colour deviation of Delta E 1.5-2.5; a crease or mark below 10 mm; and a stitch skip below 5 mm. Anything outside these bands becomes a major. A written dictionary resolves roughly 90% of disputes at the gate, against 40-50% when the classes are left to the inspector's judgement on a finished lot.

Inspection Day: What the Inspector Actually Does

A standard soft-goods inspection has six stages and takes 6-9 hours for a 500-piece lot, charged at GBP 180-320 per man-day. Stage one is quantity verification: count the cartons and reconcile against the packing list, because a short shipment of 12 pieces on a 500-piece order is a GBP 200-240 loss that is far easier to claim at the gate than from a British warehouse three weeks later.

Stage two is packing and marking: carton strength, polybag thickness, warning text, barcode legibility and country-of-origin marking. Stage three is workmanship and appearance against the golden sample. Stage four is measurement: external dimensions, sleeve clearance, strap width, weight. Stage five is on-site functional testing: zip cycling, strap pull, seam water test, colour comparison under a D65 light box. Stage six is the report, issued within 24-48 hours with photographs of every defect found.

The buyer's job is to supply the inputs. Give the inspector the golden sample, the specification with tolerances, the defect dictionary, the compliance test report numbers and the packing requirement. Without the golden sample, the inspector compares against the specification document alone and misses the 60-70% of defects that are visual. Without the tolerances, every dimension is a judgement call on the inspection report.

On-Site Tests Worth Specifying

Five tests are cheap, fast and catch the defects that reach British customers. The zipper cycle test: 200 cycles on each opening, recording pull force and slider wear. The strap pull test: 25 kg hung from each shoulder strap for 60 seconds, inspecting the bar-tack for movement. The seam water test: 40 mm of water in the base for 30 minutes on a dry towel. The drop test: 4 kg loaded, dropped from 800 mm onto concrete, checking for seam splits. The abrasion test: 500 rubs on the base panel.

Three more are worth adding for specific claims. A colour check under a D65 light box against the golden sample, recorded as a Delta E figure rather than a comment. A coating check, where a water rating is claimed, using a spray test at 8-12 mm an hour for 20 minutes. A hardware check, where metal fittings are used, with a magnet test to confirm stainless or brass rather than plated zinc, and a nickel-release spot test at GBP 12-30 per fitting type.

Specify the sample size for destructive testing separately from the AQL sample. Destructive tests destroy the unit, so they are run on 2-5 pieces pulled outside the 50-piece AQL sample, not on the sample pieces themselves. A buyer who lets the inspector cut open an AQL sample piece has effectively reduced the sample to 49 and created an argument about whether the lot passed. Pull 3 extra pieces, mark them as destructive test units and record the results on the inspection report.

What Happens When a Lot Fails

A failed lot has four possible outcomes, and the contract should rank them. Rework and re-inspect: the factory sorts 100% of the lot, replacing or repairing defective units, at GBP 1.20-3.60 per unit and 3-7 days, then a second AQL inspection is run at the buyer's or the factory's cost depending on the clause. Accept with an allowance: the buyer takes the lot at a negotiated discount of 1.5-4% per 1% of defect rate above the AQL, which works for minor defects and badly for majors.

Reject and replace: the factory makes a new lot, at 35-50 days and full cost, which is the correct outcome for a critical defect and a commercial disaster for a seasonal British launch. Cancel and refund: the deposit is returned, which is the outcome when a compliance failure cannot be reworked. Rank them in the contract as critical equals replace or cancel, major equals rework then re-inspect, minor equals allowance.

Re-inspection has its own rules. Under ISO 2859-1 switching rules, two consecutive rejections move a supplier to tightened inspection with a larger sample and a lower accept number, which is the strongest commercial pressure available short of changing supplier. Our production team accepts rework at our cost on any lot failing AQL 2.5 (ISO 2859-1) against the approved golden sample, and issues the re-inspection report with the same defect dictionary applied to the second sample.

Inspection Cost and Who Should Pay

Third-party inspection costs GBP 180-320 per man-day in the main manufacturing regions, plus GBP 40-120 for travel and GBP 30-80 for a report with photographs. A 500-piece lot is one man-day; 1,000 pieces is one to two; 2,500 is two to three. Per unit that is GBP 0.36-0.64 at 500 pieces and GBP 0.10-0.26 at 2,500, which is 1.8-3.2% of a GBP 20 landed cost and the best-value line in the entire budget.

Allocation of cost should follow fault. The buyer pays for the first inspection, which is a check on their own product. If the lot fails, the factory pays for the re-inspection, and usually for the rework at GBP 1.20-3.60 per unit. If the lot fails a second time, the factory pays for re-inspection and the buyer has the right to cancel with a deposit refund. Write those three sentences into the contract and 90% of the arguments disappear.

In-house versus third-party is a real choice. A factory's own QC department is competent and free, but it reports to the factory. A third-party inspector costs GBP 180-320 and reports to you. The hybrid is best: the factory's inline QC catches defects during production, and your third-party inspector verifies the finished lot. Ask the factory for their inline QC records - most keep them - and read them alongside the final inspection report.

Building a Quality Plan Across Multiple Orders

Quality improves across orders if the data is kept. Record the defect rate by class for every inspection: majors as a percentage, minors as a percentage, and the top three defect types by frequency. A first order typically shows 0.6-2.2% majors and 2.5-6% minors; a well-run third order should show 0.2-0.8% and 1.0-2.5%. If the numbers do not improve across three orders, the specification or the supplier is wrong.

Feed the data back into the tech pack. If print misregistration appears in every inspection, tighten the tolerance from 3 mm to 1.5 mm and pay the GBP 0.10-0.35 per unit for a jig. If zipper failures appear, move from an unspecified #5 to a branded #8 at GBP 0.45-1.20 per unit. If seam leakage appears, specify taped seams at GBP 1.20-2.80 per unit. Each fix costs less than the returns it prevents, and each one is a permanent improvement to the product specification.

Set a review cadence. Refresh physical testing on one bulk lot a year at GBP 180-420, refresh restricted-substance screening every 12-24 months at GBP 220-600 per colourway, and re-approve a golden sample whenever fabric, trim or print changes. British retailers and marketplaces increasingly require a test report dated within 24 months, and a report older than that is treated as no report. Our vetted partner facilities keep inline QC records and re-issue compliance documentation on every repeat bulk order into Great Britain.

Reading the Inspection Report and Acting on It

The report is the deliverable, not the visit. A proper soft-goods report runs 8-20 pages and contains quantity reconciliation against the packing list, carton and packing results, workmanship findings with photographs, measured dimensions and weights for 5-10 units, on-site test results with numbers, a defect count by class against the accept and reject figures, and a clear pass or fail. Expect it within 24-48 hours. Do not release the final payment without the pass page in front of you.

Read the numbers, not the summary. A report that says 'minor colour deviation observed' is not actionable; one that says 'Delta E 2.8 measured on 4 of 50 units against a tolerance of 1.5, classified major, lot fails at 4 against accept 3' is. Insist that every finding carries a measurement, a count and a classification, and that the classification follows your written defect dictionary rather than the inspector's instinct on the day.

Then act on the top three defect types. Over three orders, a British buyer typically sees the same two or three defects repeat until the specification changes: print registration until a jig is specified at GBP 0.10-0.35 per unit, zipper stiffness until a branded slider replaces a generic one at GBP 0.45-1.20, or seam leakage until taped seams are specified at GBP 1.20-2.80. Each fix costs 1.5-11% of unit price and removes 60-85% of the recurring defect rate on a repeat bulk order. Programme terms around the inspection are MOQ 500 pcs per colourway, sampling 6-10 working days and bulk production 35-50 days.