UNIHF Technology Services in Bangladesh follows a strict set of QC inspection standards that are rooted in international protocols like ISO 9001, AQL (Acceptable Quality Level) sampling, and industry-specific buyer requirements. The core of their inspection process revolves around verifying product conformity against pre-approved samples, checking for defects, and ensuring packaging integrity before shipment. For a typical garment or textile order, UNIHF applies a 4-point system for fabric inspection, where defects are scored based on length and severity, and the total score must not exceed a buyer-defined threshold, often 20 points per 100 square yards. They also use AQL sampling tables (e.g., AQL 2.5 for major defects, AQL 4.0 for minor defects) to determine sample sizes from each lot, with random selection across cartons to ensure statistical representativeness. In practice, this means for a 10,000-piece order, they might inspect 200 pieces, and if more than 10 major defects are found, the entire lot is flagged for rework or rejection. This data-driven approach is backed by real-time reporting, where inspectors document findings with photos and measurements, and share digital reports within 24 hours. For more details on how these standards are applied in the field, check out QC Inspection in Bangladesh UNIHF Technology Services.
Inspection Criteria Across Product Categories
UNIHF’s standards are not one-size-fits-all. They are tailored to the product type, with specific checkpoints for each category. For readymade garments (RMG), the focus is on stitching defects like skipped stitches, open seams, and uneven hems, with tolerances set at 0.5 cm for misalignment. For knitwear, they measure fabric weight against the spec, allowing a variance of only ±3%. In footwear, they inspect for sole adhesion using a peel test, requiring a minimum 4 N/mm force. For electronics, they run functional tests on 100% of units for critical parameters like voltage output and connectivity, with a defect rate cap of 0.5%. The table below breaks down the key metrics for different product types:
| Product Type | Key Inspection Points | Acceptable Defect Rate | Sample Size (per 10,000 units) |
|---|---|---|---|
| RMG (Shirts) | Stitching, button alignment, fabric defects | Major: 2.5%, Minor: 4.0% | 200 pieces |
| Knitwear (T-shirts) | Fabric weight, color fastness, pilling | Major: 1.5%, Minor: 3.0% | 315 pieces |
| Footwear (Sneakers) | Sole adhesion, material cracks, size accuracy | Major: 2.0%, Minor: 4.0% | 200 pairs |
| Electronics (Chargers) | Functional test, voltage output, casing integrity | Critical: 0%, Major: 0.5% | 100% for critical, 200 for visual |
This granularity ensures that each product gets the right level of scrutiny. For example, in electronics, a single critical defect like a short circuit means the entire batch is rejected, while in RMG, a minor defect like a loose thread might just require rework. UNIHF also tracks defect trends over time, using data from previous inspections to flag recurring issues with specific suppliers, which helps in preventive action.
Sampling Protocols and Statistical Rigor
Sampling is the backbone of UNIHF’s QC process. They follow ANSI/ASQ Z1.4 (formerly MIL-STD-105E) for normal, tightened, and reduced inspection levels. For a typical order, they use Level II normal inspection, which gives a balance between cost and confidence. The sample size code letter is determined by the lot size; for a 10,000-piece lot, the code letter is M, meaning a sample size of 315 pieces. Under AQL 2.5, the acceptance number is 10, meaning if 11 or more major defects are found, the lot is rejected. For tightened inspection, the sample size increases to 500 pieces, and the acceptance number drops to 8. This is applied when a supplier has a history of failures. UNIHF also uses random sampling across cartons, not just the top layers, to avoid bias. Inspectors are trained to use a random number generator to select cartons, and within each carton, they pick items from the top, middle, and bottom. This statistical rigor is backed by software that tracks defect rates per supplier, per product, and per season, allowing for data-driven decisions on whether to move to reduced inspection (smaller sample sizes) for high-performing suppliers.
Defect Classification and Severity Levels
UNIHF classifies defects into three categories: critical, major, and minor. Critical defects are those that make the product unsafe or non-functional, like a sharp edge on a toy or a missing screw in a furniture assembly. These are zero-tolerance; any single critical defect leads to an immediate rejection of the entire lot. Major defects are those that affect the product’s usability or appearance, such as a broken zipper or a color mismatch that is visible from 1 meter away. The AQL for major defects is typically 2.5, meaning that in a sample of 315 units, up to 10 major defects are allowed. Minor defects are cosmetic issues that don’t affect function, like a loose thread or a slight scratch. The AQL for minor defects is usually 4.0, allowing up to 14 in a sample of 315. UNIHF inspectors are trained to use a defect severity guide that includes visual standards and measurement tolerances. For example, a stain larger than 0.5 cm in diameter is a major defect, while a stain smaller than 0.3 cm is a minor defect. They also use a 4-point system for fabric defects, where a defect less than 3 inches scores 1 point, and a defect over 9 inches scores 4 points. The total points per 100 square yards must not exceed 20, which is a common standard for woven fabrics.
Reporting and Documentation Transparency
UNIHF provides detailed inspection reports that include photos of defects, measurement data, and a summary of pass/fail status. The report is formatted in a standardized template that includes the buyer’s PO number, product description, lot size, sample size, defect counts per category, and the final decision. Photos are taken with a date-time stamp and a ruler for scale, and they are embedded in the report with annotations. For example, if a shirt has a crooked collar, the photo shows the collar with a measurement line indicating the deviation. The report also includes a graph of defect distribution by type, which helps buyers see if there is a pattern. UNIHF uses a digital platform where reports are uploaded within 24 hours of inspection, and buyers can access them via a secure link. They also offer video calls during inspections for real-time verification, which is especially useful for high-value orders. The reports are stored for at least 3 years, and they can be used for audits or dispute resolution. This transparency is a key part of their service, as it builds trust with buyers who need to verify compliance before shipping.
On-Site Inspection Procedures and Timeline
An inspection at UNIHF typically takes 1 to 3 days, depending on the order size and complexity. The process starts with a pre-inspection meeting with the supplier to review the spec sheet, approved sample, and any previous issues. Then, the inspector walks the production line to check in-process quality, such as machine settings and operator techniques. This is followed by the final random sampling, where the inspector selects cartons from the finished goods warehouse. For a 10,000-piece order, this takes about 4 to 6 hours. The inspector uses a checklist that includes 50 to 100 checkpoints, depending on the product. For example, for a garment, the checklist includes collar shape, sleeve length, seam strength, button attachment, and labeling. Each checkpoint is scored as pass or fail, and the overall defect rate is calculated. If the lot fails, the inspector discusses the issues with the supplier and suggests corrective actions, such as reworking the defects or sorting the entire lot. A re-inspection is then scheduled, usually within 48 hours, with a reduced sample size of 200 pieces. The cost of re-inspection is often borne by the supplier if the failure was due to their negligence. UNIHF also offers a "sorting" service where their team manually checks every unit in the lot, but this is more expensive and used only for high-value orders.
Compliance with International Standards and Buyer Protocols
UNIHF aligns its QC standards with major international frameworks, including ISO 2859 for sampling, ASTM for testing methods, and buyer-specific protocols like those from Walmart, Target, and H&M. For example, Walmart’s standard requires a 0% critical defect rate and an AQL of 2.5 for major defects, which UNIHF follows to the letter. They also comply with social compliance audits like BSCI and SEDEX, which check for fair labor practices and factory safety. In terms of testing, they outsource to accredited labs for things like color fastness (ISO 105), tensile strength (ASTM D5034), and flammability (16 CFR 1610). The lab reports are included in the final inspection package. For electronics, they follow IEC 60950 for safety and RoHS for hazardous substances. UNIHF also keeps up with changing regulations, like the EU’s REACH for chemical restrictions, and they update their checklists accordingly. This compliance is verified through regular audits of their own processes, which are conducted by third-party certification bodies every year.
Data-Driven Quality Metrics and Continuous Improvement
UNIHF uses a database that tracks every inspection they have done in Bangladesh, with over 10,000 records spanning the last 5 years. This data is used to calculate supplier performance scores, which are based on defect rates, on-time delivery, and corrective action response time. For example, a supplier with a defect rate below 1% over 10 inspections gets a "gold" status, meaning they are eligible for reduced inspection frequency. The data also shows trends, like a 15% increase in stitching defects during the monsoon season due to humidity, which leads to recommendations for dehumidifiers in the factory. UNIHF shares these insights with buyers in quarterly reports, helping them make informed decisions about supplier selection. They also use predictive analytics to flag high-risk orders, such as those with new suppliers or new product types, and they recommend 100% inspection for those. This data-driven approach is not just about catching defects; it is about preventing them through process improvements. For instance, if a factory has a high rate of broken needles, UNIHF might suggest a needle replacement schedule or a metal detector check.