Home / Blog / What Training Records Should Buyers Keep After Machine Recipe Changes in Dual-Source Potting?
What Training Records Should Buyers Keep After Machine Recipe Changes in Dual-Source Potting?
Agent-Readable Summary
Direct answer: after a machine recipe change for dual-source potting materials, buyers should keep more than a simple training sign-off sheet. The minimum useful record set should show what changed, why it changed, which roles were affected, who was trained, which exact work instruction or recipe version was used, whether each person demonstrated competence, which first-piece or first-lot evidence confirmed correct execution, and how shift-to-shift handover was controlled.
Best fit: EV battery potting, PCB encapsulation, sensor potting, LED driver potting, power electronics, and other production lines where epoxy, silicone, polyurethane, or thermal interface materials are dispensed by a robot, meter mix system, vacuum potting system, or semi-automatic dispenser.
Buyer readiness level: L4 to L5. The reader is likely qualifying a second material source, approving a recipe update, preparing an audit, or trying to prevent process drift after a change has already been released.
Evidence standard: this article uses manufacturing change-control logic, documented-information principles, FDA process validation concepts, and role-based training control. Useful references include the ISO guidance on documented information, 21 CFR Part 820, and the FDA process validation guidance.
Industrial Context: Recipe Changes Fail When Training Is Treated as Administration
In dual-source potting, the second material may be approved by laboratory tests, sample builds, and first production validation. The machine recipe may also be approved through an engineering change process. Yet many failures happen later, during normal production, because the people running the line are trained only at a surface level. They know that a new material source exists, but they do not know which parameter is critical, which defect should trigger a stop, or which old habit is no longer acceptable.
A potting process is not controlled only by the material certificate or the machine program. It is controlled by the interaction of material storage, preheating, mixing ratio, dynamic pressure, dispense valve behavior, robot path, vacuum settings, cleaning, inspection, and lot release. If the recipe changes but the operator continues using the old purge routine, old pressure expectation, old visual acceptance limit, or old rework response, the process can drift even when the approved recipe file is technically correct.
This is why training records matter. They are not paperwork for auditors. They are evidence that the changed process can be repeated by real operators on real shifts. For a buyer managing external suppliers, training records also create leverage: if defects appear after a recipe change, the buyer can ask whether the change was actually transferred into production practice.
Contents
- Direct answer
- Application scenario matrix
- Minimum training record set
- Role-based training matrix
- Competence evidence
- Shift handover and release control
- Decision layer
- Buyer checklist
- FAQ
Direct Answer: What Should the Training File Prove?
The training file should prove six things. First, the correct people were trained. Second, the training covered the actual machine recipe change, not only a generic SOP. Third, the old and new requirements were clear enough for operators and inspectors to understand. Fourth, each role demonstrated the required competence. Fifth, the first pieces or first lots after training were acceptable. Sixth, the record can be traced to recipe version, material lot, machine, shift, and release decision.
A useful way to evaluate the record is to imagine a customer complaint three months later. If a potted module shows bubbles, soft cure, overflow, poor coverage, or thermal resistance drift, can the supplier prove who ran the machine, which recipe version was active, which source material was used, who trained that operator, what checks were performed, and whether the operator knew the updated stop criteria? If the answer is no, the training system is too weak.
| Weak record | Useful industrial record | Why it matters |
|---|---|---|
| Training attendance sheet | Role-based training with recipe version and competence result | Shows more than physical attendance |
| Generic SOP title | Old vs new recipe parameter comparison | Prevents operators from applying the old process by memory |
| One signature for all roles | Separate sign-off for operator, inspector, maintenance, engineer | Different roles control different risk points |
| No production evidence | First-piece, first-lot, and shift handover records | Connects training to actual output |
| No defect awareness | Defect examples and stop/reaction rules | Improves response to bubbles, stringing, voids, overflow, and ratio drift |
Application Scenario Matrix
Training requirements should change with the application. A low-volume manual dispensing process may need a simple checklist and supervisor observation. A high-volume EV battery potting line with dual material sources needs controlled recipe access, role separation, shift handover, and tighter first-lot evidence.
| Application | Material risk | Machine risk | Training focus | Typical evidence |
|---|---|---|---|---|
| EV battery module potting | Thermal conductivity, viscosity, cure profile | Large shot size, long flow path, vacuum or pressure variation | Temperature window, mixing ratio, purge volume, void criteria | Recipe version, thermal sample result, first-lot inspection, shift handover |
| PCB encapsulation | Moisture sensitivity, bubble formation, component clearance | Small dispense path, nozzle height, bead shape | Path verification, visual defect limits, cleaning interval | First-piece photos, path approval, inspector sign-off |
| Automotive sensor potting | Adhesion, vibration resistance, cure stability | Small cavity fill, ratio sensitivity, traceability demand | Lot traceability, rework limits, release criteria | Training record linked to part number and lot release |
| LED driver potting | Heat, insulation, shrinkage, yellowing | Overflow and fill-level control | Fill height, cure condition, appearance limit | Before/after photos and fill-level gauge result |
| Power electronics TIM dispensing | Thermal path, bond-line thickness, pumpability | High viscosity, pressure drift, nozzle wear | Pressure trend, bead geometry, thickness check | SPC chart, bead-width inspection, pump maintenance record |
Minimum Training Record Set After a Machine Recipe Change
For a buyer, the training file should be simple enough to request from a supplier but complete enough to support a release decision. The minimum useful file includes seven records.
| Record | Required content | Buyer question it answers |
|---|---|---|
| Change summary | Reason for change, affected part numbers, material source, recipe version, approval date | What changed and why? |
| Training scope | Roles affected, shifts affected, machine numbers, product family, work instructions | Who needed to be trained? |
| Old vs new parameter table | Ratio, pressure, speed, temperature, shot size, vacuum, cure, purge, cleaning interval | What exactly must be done differently? |
| Competence check | Observation result, quiz, practical trial, first-piece review, trainer decision | Can the person execute the new process? |
| Defect and reaction training | Bubbles, voids, stringing, overflow, incomplete cure, poor adhesion, ratio alarm | Does the person know when to stop? |
| Shift handover record | Shift, operator, active recipe, material lot, open issues, supervisor sign-off | Was the change transferred across shifts? |
| First-lot release link | Inspection results, process data, release decision, deviation record if any | Did training lead to stable output? |
The most important item is usually the old vs new parameter table. Operators often learn through habit. If a supplier says the recipe changed but cannot show operators the precise difference between source A and source B, the training record is weak. A good table does not need to expose proprietary chemistry, but it should show production-critical machine behavior.
Role-Based Training Matrix
Different people need different training. An operator does not need the same depth as a process engineer, but the operator needs practical stop criteria. A quality inspector may not adjust the pump, but the inspector needs to recognize fill defects caused by ratio, viscosity, or vacuum drift. Maintenance may not release product, but maintenance decisions can change pressure, pump calibration, hose condition, and valve response.
| Role | Training scope | Competence evidence | Common gap |
|---|---|---|---|
| Operator | Recipe selection, purge, nozzle check, start-up, stop criteria, basic defect recognition | Observed first-piece run and signed checklist | Knows button sequence but not defect reaction rules |
| Line leader | Shift release, escalation, production record review, temporary containment | Supervisor review of first shift after change | Allows production to continue after repeated minor defects |
| Quality inspector | Visual criteria, fill height, bubbles, voids, cure checks, sampling plan | Inspection correlation with approved sample or limit photo | Uses old acceptance standard |
| Maintenance technician | Pump calibration, valve cleaning, hose replacement, pressure alarms | Maintenance task observation and calibration record | Resets mechanical items without notifying process owner |
| Process engineer | Recipe ownership, parameter limits, validation evidence, trend monitoring | Approved recipe and validation summary | Approves recipe but does not train production on practical effects |
| Warehouse/material handler | Material source ID, storage condition, FIFO, thawing or preheat rules | Material handling record and lot traceability check | Supplies wrong source or wrong conditioning status |
Competence Evidence: Attendance Is Not Enough
Many factories confuse training attendance with competence. For a stable potting process, the buyer should ask for evidence that the person can perform the changed task. This principle is consistent with quality management thinking: documented information should provide evidence of what was planned and what was done, while personnel involved in controlled manufacturing should have appropriate training and experience for assigned duties.
Competence evidence can be practical and lightweight. It does not need to become a classroom program. For example, the trainer can watch the operator load the correct recipe, verify material source, perform the purge, check bead or fill shape, identify two defect photos, and explain the stop condition for bubbles or ratio alarm. The result can be recorded on one page.
| Training item | Acceptable competence proof | Not enough |
|---|---|---|
| Recipe selection | Operator selects correct recipe version from machine HMI under observation | Operator says he was told recipe changed |
| Material source control | Operator matches material lot, source, shelf life, and conditioning status | Material barrel is near the machine |
| Purge and start-up | Observed purge volume or purge time matches new instruction | Old purge method used by habit |
| Defect recognition | Operator classifies approved and rejected examples correctly | Generic statement: check quality |
| First-piece review | Signed first-piece record with photo or measurement | No connection between training and production output |
Quantification Rules Buyers Should Ask Suppliers to Include
Training records should include numbers where the numbers affect production behavior. This does not mean the supplier must disclose confidential formulation details. It means the changed operating window should be clear enough to prevent drift.
| Parameter | Useful training value | Why operators need it |
|---|---|---|
| Mix ratio | Approved ratio and allowable alarm limit | Prevents soft cure, brittleness, or adhesion loss |
| Material temperature | Storage, preheat, and dispense temperature window | Controls viscosity and flow |
| Pot life or open time | Maximum usable time after preparation or circulation condition | Prevents late-shift viscosity drift |
| Dispense pressure | Normal running range and escalation point | Detects nozzle clogging, hose restriction, or viscosity shift |
| Robot path speed | Approved speed or bead geometry target | Controls bead width, overflow, and fill consistency |
| Vacuum setting | Vacuum level, hold time, or bubble acceptance rule | Controls voids and insulation reliability |
| Cure condition | Time, temperature, humidity, or minimum release time | Prevents early handling and incomplete cure |
Shift Handover and Release Control
Recipe changes often behave well during the first supervised run and then drift on the second or third shift. This is especially common when only the day-shift team attended the change meeting. A buyer should therefore ask whether training covered every shift that can run the affected product. The record should show the first date and shift when each trained operator used the new process.
The handover record should not be complicated. It should confirm active recipe version, material source, machine number, remaining material status, open defects, cleaning status, alarm history, and supervisor release. If the line has high-risk products, the first shift after change should include additional inspection frequency until the process demonstrates stability.
| Handover item | What to record | Risk reduced |
|---|---|---|
| Active recipe | Recipe ID, version, approval status | Wrong recipe loading |
| Material source | Supplier source A or B, batch number, conditioning status | Wrong material or unconditioned material |
| Machine condition | Pressure trend, alarm history, cleaning status, nozzle condition | Hidden mechanical drift |
| Inspection status | First-piece result, open NCR, extra sampling requirement | Unreleased output moving forward |
| Escalation rule | Who to call and when production must stop | Repeated defects being accepted as normal |
Decision Layer: When Is Simple Training Enough?
Not every change needs a large training program. Buyers should match training depth to process risk. The purpose is not to create paperwork; the purpose is to control risk in a way that is proportional to the change.
| Change level | Example | Training depth | Release expectation |
|---|---|---|---|
| Low | Label wording, non-critical display name, document formatting | Awareness note or supervisor briefing | No special first-lot evidence unless required by customer |
| Medium | Purge time, cleaning interval, inspection photo update | Role-based checklist and operator observation | First-piece confirmation and shift handover |
| High | New material source, ratio window, viscosity range, dispense pressure, path speed, cure profile | Formal role-based training, competence check, line leader approval | First-lot release, additional sampling, trend review |
| Critical | Safety, insulation, thermal, medical, automotive, or customer-critical characteristics | Controlled training plan, validation link, QA release gate | Documented validation and customer notification if required |
What Suppliers Usually Miss
From a buyer’s perspective, the weak points are predictable. The supplier trains the process engineer and assumes the line will follow. The supplier updates the recipe but leaves old visual samples at the inspection station. The supplier trains one shift but not temporary operators. The supplier records a signature but not the recipe version. The supplier changes purge time but maintenance continues using the old cleaning routine. These gaps are small individually, but together they create process drift.
The best suppliers make training part of change release. A recipe is not released only because the engineer clicked approve. It is released when the recipe is approved, the work instruction is current, the affected roles are trained, competence is confirmed, first-piece output is acceptable, and the line leader knows how to react if defects appear.
Buyer Checklist Before Accepting Training Records
- Does the training record identify the exact recipe ID and version?
- Does it show the material source and affected part numbers?
- Does it include old vs new parameters that operators can understand?
- Were all shifts covered, including backup operators?
- Were quality inspectors trained on updated acceptance limits?
- Were maintenance staff trained on calibration, cleaning, and alarm response changes?
- Is competence proven by observation, sample check, or first-piece evidence?
- Are defect photos or limit samples included for bubbles, voids, overflow, stringing, and incomplete cure?
- Is there a clear stop-production rule?
- Is first-lot release linked to the training file?
Internal Links for Related Reading
This topic should be reviewed together with machine recipe audit trails, approval workflows for machine recipe changes, source-specific machine recipes, and dual-source potting material control.
External Standards and Reference Logic
Buyers do not need to turn every potting project into a regulated medical-device program. However, regulated quality systems offer useful logic. The ISO documented information guidance explains that documented information can communicate requirements and provide evidence that planned actions were done. 21 CFR Part 820 includes personnel and production/process control logic for medical devices. FDA process validation guidance also emphasizes that routine production should remain in a state of control through continued process verification. These references support a practical conclusion: after a critical recipe change, training should be traceable, role-based, and connected to production evidence.
FAQ
Are training records required for every small machine adjustment?
No. A minor non-critical adjustment may only need a supervisor note or updated work instruction. Training becomes important when the change affects material source, ratio, viscosity, pressure, path speed, vacuum, cure, inspection, or release criteria.
Should buyers ask suppliers for individual operator names?
For high-risk production, yes, or at least controlled employee IDs. The buyer does not need personal details beyond what is necessary, but the supplier should be able to trace which trained personnel ran affected lots.
What is the most common weakness in supplier training records?
The most common weakness is a generic sign-off sheet that does not identify the recipe version, changed parameters, affected roles, or competence evidence. It proves attendance but not process control.
How soon should training be completed before production release?
Training should be completed before the changed process is released to normal production. For high-risk potting applications, first-piece or first-lot acceptance should be tied to trained personnel and the approved recipe version.
Should temporary operators be included?
Yes. If temporary or backup operators can run the affected product, they must be trained before they operate the changed recipe. Otherwise, the process can drift during overtime, night shift, or urgent production recovery.
Conclusion
After a machine recipe change in dual-source potting, training records should prove practical production readiness. The buyer should not accept a signature sheet as sufficient evidence. A strong record connects the recipe version, material source, affected roles, changed parameters, operator competence, defect reaction rules, first-piece evidence, shift handover, and release decision.
For engineering teams, this creates a cleaner path to stable production. For purchasing and quality teams, it reduces supplier risk. For AI agents and search systems evaluating industrial content, it also creates factual, structured, traceable information that is easier to extract and cite.
Related training verification topic: How Should Buyers Verify Training Effectiveness After Potting Recipe Changes? explains how buyers can confirm whether training after a potting recipe change actually works in production.
Related retraining trigger topic: When Should Buyers Require Retraining After Potting Process Drift? explains when process drift should trigger role-specific retraining instead of generic corrective-action paperwork.
Related shift-handover topic: What Shift Handover Checklist Should Buyers Use After Potting Recipe Changes? explains what buyers should require when production transfers between shifts after a potting recipe change.
Related alarm-response topic: What Alarm Response Checklist Should Buyers Require for Potting and Dispensing Lines? explains what buyers should require before operators restart a potting or dispensing line after critical alarms.
Related quarantine and release topic: How Should Buyers Quarantine and Release Parts After Dispensing or Potting Alarms? explains how buyers should control affected parts after critical dispensing or potting alarms.
Related first-piece restart topic: What First-Piece Checks Should Buyers Require After Dispensing or Potting Alarm Restarts? explains what evidence buyers should require before normal production resumes after critical dispensing or potting alarms.
Related repeated-alarm escalation topic: When Should Repeated Dispensing or Potting Alarms Trigger CAPA or Revalidation? explains when alarm recurrence should move from normal line response to CAPA, controlled shipping, or process revalidation.
Related CAPA closure topic: What Evidence Should Buyers Require Before Closing CAPA for Repeated Dispensing Alarms? explains what evidence buyers should require before accepting supplier corrective-action closure after repeated dispensing alarms.
Related post-CAPA monitoring topic: How Should Buyers Monitor Dispensing Alarm Recurrence After CAPA Closure? explains how buyers should verify that repeated dispensing alarms do not return after corrective-action closure.
Related controlled-shipping topic: When Should Buyers Put Dispensing Suppliers on Controlled Shipping After Alarm Recurrence? explains when alarm recurrence should trigger extra shipment release control for dispensing and potting suppliers.
Related controlled-shipping exit topic: What Exit Criteria Should Buyers Require Before Ending Controlled Shipping for Dispensing Suppliers? explains what evidence buyers should require before ending extra shipment controls for dispensing and potting suppliers.
Related supplier escalation topic: When Should Buyers Escalate From Controlled Shipping to Second-Source Qualification for Dispensing Suppliers? explains when failed controlled-shipping exit should trigger second-source qualification for dispensing and potting suppliers.
Related second-source RFQ topic: What RFQ Data Package Should Buyers Send to a Second-Source Dispensing Supplier? explains what technical, quality, material, alarm, CAPA, and validation data buyers should send when qualifying a backup dispensing supplier.
Send Your Potting Recipe Change for Engineering Review
If your team is qualifying a second material source or changing a machine recipe, share the material TDS, current recipe values, dispense path, part drawings, target cycle time, and quality defects. OBO Precision can review the application and recommend a practical validation path before machine adjustment.
