A needle breaks while an operator is sewing a pocket. The machine stops, and two pieces of the needle are found. A third piece is missing. Meanwhile, a trolley has taken completed garments to the next section. The immediate problem is no longer confined to one machine.
Needle and sharp-tool control must answer several linked questions. What is missing? Which goods could contain it? Where are those goods now? What evidence will support the decision on those goods? A signed replacement slip answers only part of that problem.
This guide builds a control system for an apparel factory. The incident and stock figures are made-up examples. They show how records can support sound decisions. They do not set a search radius, detector limit or buyer standard. Get those rules from the product requirements, buyer agreement and approved site procedure.
Start with the product requirement and the actual tools
Find out which product-safety rules apply before writing the needle procedure. Buyers may need different records, tests and forms of approval. A rule copied from another factory can leave a gap even when staff follow it in full.
Marks & Spencer's public product-safety policy illustrates this distinction. It sets out supplier duties and how to manage incidents. It directs suppliers to controlled platforms for the full production standards. The public policy is not a complete needle-control standard. Get the current rules for the order you are making.
The factory's tool list should follow the work. Sewing needles are an obvious starting point, but hand needles, tag-gun needles, scissors, trimming blades and maintenance items can also reach goods. Consider packing and subcontract work as well as the sewing floor.
Give each type a defined issue, return and exception process. A reusable pair of scissors needs an individual identity. For a needle, record its exact type and each change. For a blade assembly, record new parts as well as the whole tool.
Do not assume one detection method covers this entire list. The material, shape and location of a missing item matter. First reduce the risk and keep track of tools. Inspection adds evidence, but each method has limits.
Separate the three decisions after a breakage
Three decisions can become confused: making the machine safe, allowing work to restart, and releasing held goods. Some evidence may serve all three. Each decision still needs its own approval.
First, stop the work and hold goods that may contain a fragment. A trained person should make the machine safe under its operating and isolation procedure. Do this before searching near moving parts or removing a needle. The search must not put anyone at risk.
As a jurisdiction-specific example, US OSHA's machine-guarding rule requires protection from relevant machine hazards. Use the rules and machine instructions that apply at your site. A production target cannot override them.
Next, find out what the machine needs before it can restart. Check the needle system, how it was fitted, the settings and any damaged parts as needed. A trained maintenance or production person records the work and checks that the machine is safe to use.
Finally, a quality person with the right authority decides what happens to the goods. The machine may restart once its fault is fixed while an earlier batch stays on hold. Finding a fragment does not, on its own, prove that the machine is ready to run.
Give the incident record separate fields for machine restart and the decision on goods. A signature in one box must not become permission for every next step.
Build a hold around evidence of exposure
The nearest bundle is not automatically the full affected quantity. Goods may have moved since the last reliable check, and a fragment may have reached nearby work. Base the first hold on what you know about movement, timing and the breakage.
In this example, the supervisor finds goods in four places linked to the work under review. Each row counts whole garments. No garment is counted twice.
Location at the time of the hold | Garments identified | Immediate record |
|---|---|---|
At the affected workstation | 6 | Tray identity and operator's record |
Work trolley awaiting the next operation | 54 | Trolley and bundle references |
Downstream inspection queue | 36 | Transfer record and physical count |
Finishing trolley | 24 | Receiving record and physical count |
Total initial scope | 120 | One incident reference across all locations |
The count is 6 + 54 + 36 + 24 = 120 garments. That is the total in this example; it is not a fixed rule for every breakage. The supervisor finds eight nearby garments that may also have been exposed. These join the hold, so the current count is 128.
Write down why each group was included. If later evidence changes the scope, keep the old count, the reason and the approval. Changing 128 to a smaller figure without a note loses the history of the decision.
The physical hold and the system status must agree. A red label on a trolley is weak protection if its contents can still be issued for packing. Likewise, a computer hold is incomplete if an unmarked trolley can move without a scan.
Use the same incident reference on goods, movement records and the release decision. The ApparelDiary guide to rework loops is useful when held goods enter a separate inspection or repair route. Keep each group linked to its source as it moves through that route. Loose pieces must not rejoin normal work without that link.
Reconstruct the needle without creating another hazard
The search and the evidence record serve different purposes. A search tries to find missing material. The record shows what was found. It helps the reviewer judge whether the needle is complete.
Use the approved collection method, suitable tools and a secure container. Do not leave fragments loose on a desk or carry them in a pocket. Train the person to collect pieces safely and to use the approved comparison method. They must also know what that method cannot prove.
Groz-Beckert's Ideal Needle Handling documentation provides a relevant manufacturer example. Its system links needle changes to records and images. A return box helps keep fragments secure during review. This shows how safe handling and records can work together. The buyer's approval rules still apply.
Record the needle system and size, the machine, the time of the event and the recovered pieces. A comparison should use the correct needle reference. A needle that looks much the same may have a different length or shape. Do not dismiss an unexplained gap as a drawing error.
A photograph needs context. Include the case number and a suitable scale where the method calls for one. Keep the original file. A close-up may be hard to read later if it lacks the needle type, scale or case link.
Avoid marking a case complete merely because the expected number of pieces was found. A needle can break into a different number of fragments, and one piece may belong to an earlier event. Judge whether the object is complete from the pieces and the evidence. There is no standard count of fragments that proves this.
If completeness remains uncertain, the incident stays open. Pass it to the next person in the approved review route. Stronger wording does not fill the gaps in the evidence.
Know what the detector is designed to find
The terms needle detector and metal detector are often used loosely. Check the actual equipment specification before using either term in a release procedure.
Hashima explains its needle-detection principle in terms of a response to magnetic material. CEIA describes its THS/MN21 textile system as detecting ferrous and magnetic contaminants. Neither description means that the system will find every metal, blade, pin or other foreign object.
Write down how the chosen system addresses each likely hazard. Ask its maker about the needle material, the finished garment and the test setup. If the method cannot address a hazard, the quality plan needs another approved control. In some cases, the goods may need a different route for review.
Intended metal parts also matter. Zips, studs and other trims may affect the test. A trim may be sold as suitable for needle detection. Still check it in the finished garment and the planned setup. Several trims together may behave differently from one tested alone.
A detector pass means the product did not trigger the system under those test conditions. It does not prove that no fragment of any size or material remains. This limit matters most after a known breakage when part of the needle is still missing.
Use all the evidence to decide the case. This includes the missing object, found pieces, possible exposure and records that trace the goods. It also includes results from a suitable test. A routine screening result must not replace the search and review for a known missing fragment.
Write a test method that can be repeated
A sensitivity setting on a display is not, by itself, a performance check. The test must state what is checked and what response counts as a pass.
Hashima's guidance on test cards and pieces describes iron-sphere references and ways to vary their height. This shows why the test-piece type and position matter. It does not set one sensitivity rule for every factory or garment.
Agree the method with the machine supplier and check it against the buyer's rules. Record the test-piece ID, material and stated size. Add the machine ID and garment setup. A worn label or unknown test card cannot support a sound safety claim.
Define the test positions, how goods pass through, the settings and the required response. State when checks are due and what calls for an extra check. The plan may require checks at start-up or after a change, move or unexplained fault. Write the actual schedule for that system.
Test the whole action that protects the flow. If the approved response includes a belt stop, alarm or rejection step, observing a light alone is not enough. The operator must know how to spot a failed check and stop goods from moving on.
Record actual results rather than filling a shift's boxes in advance. When a check fails, isolate the relevant product interval and investigate the equipment. The last valid check and the movement records help define that interval. The approved team must review the event; it cannot assume a fixed quantity.
Once the fault is fixed, successful checks support a return to use. Goods that passed through during the uncertain period still need their own decision. Link that decision to the evidence and approval.
Investigate alarms without weakening the control
Repeated alarms can create pressure to lower sensitivity or wave products through. Treat that pressure as a sign that the inspection process needs attention.
Hashima's troubleshooting guidance identifies belt contamination and external interference as possible causes of unwanted responses. It also describes equipment-specific conditions affecting product sensing. Get trained support to check these causes. They do not justify ignoring an alarm.
Separate product, environment and equipment questions. Does the response follow one garment? Does it recur at the same belt position? Did a trolley, fan or nearby machine move? Record what you see. Check a possible cause before treating it as the reason for the alarm.
Hashima's installation guidance also identifies nearby equipment and moving metal as concerns for its detectors. Moving a machine to a spare corner may call for a new site check. Record the approved place and setup as part of the test method.
A setting change needs control. Record who approved it, why it was needed and which checks confirmed the new setup. A lower setting may speed the flow but fail the required test. If so, the machine is not ready for that product.
Keep suspect goods clearly marked throughout this work. Passing the same garment through several times until one pass gives no alarm is not an investigation. Explain the result and decide what happens to the goods under the approved method.
Reconcile new needles, returns and open incidents
A stock balance can reveal uncontrolled access even when no breakage has been reported. Keep it simple enough to complete during the shift. Another person should be able to check it without guessing what the entries mean.
Suppose a secure store starts with 200 unused needles. It receives 100 more, issues 72 and returns two unused needles to the supplier. Its expected unused stock is:
Expected closing stock = 200 + 100 − 72 − 2 = 226 needles.
If the physical count is 225, the difference is one needle. Check the timing, unit of issue, receipt, supplier return and any correction records. Do not invent an issue to make the balance agree.
The issue ledger and the used-needle record answer separate questions. In this example, the 72 changes comprise 60 worn-needle changes, nine planned changes and three breakage events. That gives 72 change events. It does not show whether all the pieces of each broken needle were found.
Record | What it proves when supported | What it does not prove |
|---|---|---|
Unused stock balance | Accounted movements of new needles | Completeness of a broken needle |
Change record | A specified replacement at a machine | Safe disposition of exposed garments |
Fragment record | What material was recovered and assessed | Correct detector performance |
Detector test record | The observed response under a defined test | Absence of every possible contaminant |
Product disposition | The authorised outcome for identified goods | Elimination of the original machine fault |
An open incident must remain visible after a new needle is issued. Otherwise, an efficient store transaction can hide an unresolved product hold. Review the stock difference and the breakage case through their own routes. Link them where the evidence shows a connection.
Extend the same discipline to other sharp tools
Scissors and trimming tools need a clear owner or station, a storage position and a handover rule. The record should show when a tool changes location, is replaced or is removed for repair. A tally without tool IDs is weak evidence when several similar tools move around.
A tether can help prevent a tool being misplaced where it is suitable for the task. Check its design and position for snagging, reach and safe movement. The same tie will not suit every tool and task.
Maintenance creates a different route for loose items. Before handing the station back, the technician should check all tools, old parts and new parts. If a broken part is incomplete, assess where the missing material could have gone. A routine repair sign-off does not close that question.
Packing areas should control tools and fasteners that might enter an open garment or carton. The appropriate method depends on the product and customer requirements. Do not use a banned item just to hold temporary paperwork.
Include outsourced work in the plan. Agree the subcontractor's records and hold-notice process before goods leave. Also agree how they must state the return status. A correct count on return does not prove that sharp-tool controls were followed off site. The return papers should name any incident and record its outcome.
Give product release its own evidence trail
Return to the example hold of 128 garments. Finding the missing fragment would be useful evidence. The approved reviewer must still check that it is the right piece. They must also check the hold boundary and the required tests on goods. If the fragment is still missing, record that uncertainty in the decision.
There is no sound universal instruction to destroy every affected batch, and no sound universal instruction to release it after a detector pass. Follow the rules that apply, with a trained review of the actual case. Some cases need the buyer's input or expert support.
The release record should name the goods and the evidence reviewed. Add any tests or rework and the person with authority to decide. If goods are rejected or destroyed, keep a count and proof of the action. They must not return unnoticed to saleable stock.
If goods that may be affected have left the factory, use the product-safety process at once. Keep the records needed to trace them. Involve the responsible commercial and quality staff. Duties to report and buyer actions depend on the market and the case. An informal agreement must not bypass them.
The quality assurance and quality control guide explains the wider distinction between a system and its checks. Here, the system is the controlled path from issue to incident closure. An inspection result is one piece of evidence within that path.
Test the procedure before a busy shift exposes a gap
A short desk exercise can reveal gaps without putting loose sharp objects near production. Use labelled case cards or other approved training aids. Give the team a made-up event, a movement record and a partly filled hold log. Ask them to explain their next decisions.
For example, the exercise might say that a trolley reached finishing while its transfer was still unposted. Can the team find the physical goods? Who can block the shipment? Does the night shift have access to the required customer instruction? These questions test whether the procedure works in practice.
Include an absent approver and an unavailable computer in separate trials. The backup process should keep case IDs, times and approval history. A paper record can work if it is controlled and later checked against the system. A spoken order with no record is much harder to verify.
Trace one completed practice case from the needle issue to the final garment count. Check whether each reference leads to the next record. The search log may name one machine while the issue slip names another. Fix that design problem before staff face a real incident.
The exercise should also test understanding. Ask the operator to show where held goods belong and whom to contact. Ask the reviewer what evidence is still missing. A training sheet proves the person attended. Watching them do the task gives better evidence of whether they can carry it out.
Record the gaps, assign owners and repeat only the parts that failed after they are corrected. Keep the exercise separate from real incident data. Made-up breakages must not distort the factory's trend reports.
Use the records to prevent another incident
Closing a product hold should lead to a separate review of the cause. The review may need to check the needle type, its condition and how it was used. It may also need to check the machine setup and fabric. A possible cause becomes useful when there is a way to test it.
Compare like conditions. Six breakages during 240,000 comparable sewing operations give:
Breakages per 100,000 operations = 6 ÷ 240,000 × 100,000 = 2.5.
This is a worked rate, not an acceptable limit. Another line may report three breakages. That count tells little unless you know its workload and working conditions. Keep the operation basis stable, and do not confuse change events with individual recovered fragments.
Track the age of open cases, unexplained stock gaps, missed tests and repeat causes. A low breakage count is not success if staff fear reporting an event. Check that the process lets them report promptly and hold goods safely.
Try one controlled change and record it. Observe similar work to see whether the problem returns. Check for changes in quality and safe working conditions too. A downtime saving has little value if it depends on weaker checks or holds that go unrecorded.
At handover, the next team should know which goods remain held and which machines are ready. Each open action needs an owner. The records should let another trained person make the next correct decision without guessing.
Discussion0
Loading the discussion…