QUICK ANSWER
Contract assembly is not simply outsourced labor. A reliable program converts an engineering definition—BOM, drawings, specifications, approved components and acceptance criteria—into a controlled production system that can repeatedly build, test, document and release the correct configuration.
FABRIOX ASSEMBLY FRAMEWORK |
1. What Is Contract Assembly?
Contract assembly is the outsourced integration of components, fabricated parts, purchased items, electrical content and subassemblies into a higher-level product or module. The scope can range from mechanical fastening and kitting to electromechanical integration, wiring, functional testing, labeling, documentation and shipment-ready packaging.
The commercial attraction is obvious: the OEM can reduce internal touch labor, supplier interfaces, work-in-process and line-side complexity. The engineering challenge is equally important: once multiple parts become one assembly, errors can propagate across configuration, fit, torque, wiring, software, testing and traceability.
2. The Assembly Supplier Becomes a System Integrator
A component supplier can sometimes release one drawing-controlled part. An assembly partner must coordinate many inputs and produce one verified output. That changes the operating model.
Component manufacturing | Contract assembly / integration |
|---|---|
One or few drawings | BOM + multiple drawings + specifications + variants |
Local process capability | Cross-process and supplier coordination |
Part-level inspection | Interface + configuration + functional verification |
Material traceability | Component + lot + serial + revision traceability |
One manufacturing route | Kitting, sequence, fastening, wiring, test, pack |
Part nonconformance | Potential system-level failure |
CORE PRINCIPLE
THE ASSEMBLY LINE IS WHERE SUPPLY-CHAIN VARIATION BECOMES PRODUCT BEHAVIOR.
3. Start With a Controlled Product Definition
No assembly process can be more stable than the engineering definition it receives. The production package should identify the released BOM, drawings, specifications, software/firmware where applicable, approved substitutes, labels, packaging and acceptance criteria.
Unique part numbers and descriptions
Revision-controlled drawings/specifications
Quantity per assembly
Approved manufacturer/supplier where required
Variant/options logic
Special characteristics / CTQs
Fastener and torque requirements
Test specification and limits
Label/serial-number rules
Packaging configuration
CONFIGURATION RULE
DO NOT BUILD FROM A COLLECTION OF EMAILS. Build from one released configuration baseline.
4. BOM Integrity Is a Quality Characteristic
A physically perfect assembly built with the wrong revision, connector, seal, fastener grade or software configuration is still nonconforming. BOM accuracy therefore deserves the same control as dimensional quality.
Engineering changes should update the BOM, drawings, work instructions, test requirements and inventory disposition as one coordinated release.
5. Variant and Option Management
Product families often share most components but differ in harnesses, brackets, software, labels, hoses, sensors or customer-specific options. This creates a high-risk 'almost the same' environment.
Control | Purpose |
|---|---|
Variant matrix | Makes configuration differences explicit |
Unique kit / traveler | Prevents cross-mixing |
Barcode / scan verification | Confirms selected component |
Physical segregation | Reduces wrong-part selection |
Poka-yoke connectors / fixtures | Prevents impossible combinations |
Final configuration check | Confirms released variant before shipment |
6. Supplier Inputs: Buy the Assembly, Manage the Network
Contract assembly often includes buyer-nominated components, assembler-sourced parts and custom manufactured content. Responsibility must be explicit for sourcing, supplier qualification, incoming inspection, shortages, obsolescence and substitutions.
The assembly partner should not silently substitute a connector, fastener, adhesive, cable or bearing because the original is unavailable. Functional equivalence is an engineering decision unless the released specification explicitly permits alternatives.
SUBSTITUTION RULE
NO UNCONTROLLED SUBSTITUTION. Commercial urgency does not override configuration control.
7. Incoming Quality Should Be Risk-Based
Inspecting every characteristic on every incoming component is usually inefficient. Inspecting nothing assumes upstream processes are perfect. A risk-based plan considers supplier performance, criticality, detectability at later stages, certificate requirements and failure consequence.
Input category | Typical control |
|---|---|
Critical custom part | Dimensional / material / certificate verification |
Standard fastener | Approved source, grade/coating identity, lot control |
Electrical component | Part number / revision / ESD and storage controls as applicable |
Seal / adhesive / chemical | Shelf life, batch, storage condition |
Purchased module | Identity + defined functional/visual check |
Stable low-risk commodity | Reduced inspection based on performance |
8. Nonconforming Inputs Need Containment Before They Reach the Line
Incoming discrepancies should be identified, segregated and dispositioned before kitting. 'Use as is', rework, concession and deviation require defined authority.
A mature assembly process does not let operators discover supplier defects after the product is half assembled.
9. Kitting Is a Production-Control Process
Kitting converts a complex BOM into a controlled set of inputs for one assembly, batch or work order. Good kitting reduces search time, shortages, mixed revisions and line-side inventory.
Kit by work order / serial number where appropriate.
Separate visually similar variants.
Verify quantity and revision before line release.
Control partial kits and shortages explicitly.
Keep traceability labels linked to the kit.
Do not allow unrecorded line-side substitution.
KITTING PRINCIPLE
THE CHEAPEST ASSEMBLY ERROR IS THE ONE PREVENTED BEFORE THE FIRST FASTENER IS INSTALLED.
10. Work Instructions Must Translate Engineering Into Repeatable Work
A drawing defines the product; it rarely defines the complete assembly sequence. Work instructions should translate design intent into operator-executable steps.
Work-instruction element | Example |
|---|---|
Sequence | Install seal before enclosure closure |
Orientation | Connector key faces datum side |
Tool | Calibrated electric nutrunning spindle |
Parameter | Torque / angle / press force / cure time |
Consumable | Approved threadlocker / grease / adhesive |
Inspection point | Witness mark / gap / connector lock |
Traceability | Record tool result against serial number |
Safety / handling | Lift point, ESD, clean surface |
Visual instructions, photos and controlled diagrams are useful when they clarify orientation or routing, but they must be revision-controlled just like text.
11. Standard Work vs Tribal Knowledge
If product quality depends on one experienced operator remembering a special sequence, the process is fragile. The objective is not to remove skill; it is to capture critical knowledge in the controlled production system.
STANDARD-WORK PRINCIPLE
EXPERTISE SHOULD IMPROVE THE PROCESS—not remain hidden inside one person's memory.
12. Sequence Matters
Assembly order affects access, alignment, damage risk, torque, cable routing and testability. A fastener installed too early may block a connector; a cover closed too soon may hide a verification point.
Sequence should be designed together with fixtures, tooling, test points and rework strategy.
13. Mechanical Fastening: Torque Is a Process, Not a Number
A torque value on a drawing does not by itself create a controlled bolted joint. Joint friction, fastener grade/coating, lubrication, prevailing torque, washer stack, thread condition, tool accuracy and tightening strategy all influence achieved preload.
Critical joints may require controlled torque, torque-angle, yield-based strategies or customer-specific methods. The appropriate method depends on joint function and engineering requirements.
FASTENING RULE
CONTROL THE JOINT SYSTEM—not only the wrench setting.
14. Torque Tool Control
Use tools suitable for the required range and joint type.
Maintain calibration/verification status.
Prevent use of expired or incorrect programs.
Link tightening program to product variant where possible.
Define reaction strategy and operator ergonomics.
Record results for critical joints when required.
Control sockets, extensions and adapters that affect application.
A traceable torque record is valuable only when the tool, program, fastener and serial number relationship is reliable.
15. Torque Marking Is Not Torque Verification
A paint witness mark can show that a fastener passed through a marking step or help detect later movement. It does not prove that the required preload was achieved unless the marking process is tied to a controlled tightening operation.
EVIDENCE PRINCIPLE
A VISUAL MARK IS PROCESS EVIDENCE—not a substitute for a validated tightening result.
16. Press Fits and Force-Displacement Processes
Bearings, bushings, pins and inserts may require press assembly. For critical interfaces, force and displacement monitoring can detect missing parts, wrong orientation, interference variation or incomplete seating.
The acceptance window should be established from engineering and process validation, not invented after production data appears.
17. Adhesives, Sealants and Threadlockers
Bonding processes depend on material identity, shelf life, storage, surface preparation, mix ratio where applicable, application amount, open time, cure time and temperature.
A bead that 'looks right' may not provide a controlled seal. Critical adhesive/sealant processes should define the material, preparation, application and cure conditions.
CONSUMABLE RULE
TRACE THE MATERIAL LOT AND CONTROL THE PROCESS WINDOW when the consumable creates function.
18. Lubrication and Grease
Lubrication can be functional, assembly-aiding or both. Wrong grease, wrong quantity or contamination can affect seals, bearings, torque and long-term reliability.
Specify location, material, amount/range and contamination controls where lubrication is critical.
19. Wiring Harness Integration
Harness integration introduces polarity, connector locking, routing, bend radius, strain relief, chafing, separation, clipping and continuity risks. If harness manufacture is included, acceptance should use the contractually applicable revision/class of IPC/WHMA-A-620 or another specified standard.
As of August 2026, IPC/WHMA-A-620F-2025 is the current revision identified by ANSI's standards store; it describes requirements and acceptance for crimped, mechanically secured and soldered interconnections and related cable/harness assembly activities.
STANDARDS GUARDRAIL
A STANDARD NAME IS NOT A COMPLETE REQUIREMENT. Define the applicable revision, class/addendum where relevant, customer deviations and inspection/test scope.
20. Harness Installation Is Different From Harness Manufacture
A conforming harness can still fail in the final product if it is routed against a sharp edge, overstretched, pinched, connected to the wrong port or installed without strain relief.
Verify connector identity and keying.
Confirm full latch / secondary lock.
Control routing and clip locations.
Protect against abrasion and heat sources.
Respect minimum bend / service-loop requirements from design.
Perform continuity / functional checks at the correct integration stage.
21. Electromechanical Integration
Electromechanical assemblies combine mechanical interfaces with motors, sensors, switches, actuators, power distribution, harnesses and sometimes control electronics. The integration plan should define both mechanical and electrical acceptance.
Mechanical CTQ | Electrical / functional CTQ |
|---|---|
Torque / retention | Continuity / insulation as specified |
Alignment / gap | Correct connector / pinout |
Bearing / motion | Sensor response |
Seal / enclosure | Power-up behavior |
Fastener presence | Current / voltage / signal where specified |
Label / orientation | Firmware / parameter configuration where applicable |
22. PCB and Electronic Subassemblies
If the assembly includes electronic assemblies, the applicable IPC requirements should be contractually selected rather than assumed. IPC lists IPC-A-610 for electronic-assembly acceptability and J-STD-001 for soldered electrical/electronic assemblies; revision/class/addenda should match the product and customer requirement.
The contract assembler should also define ESD controls, handling, programming and functional-test responsibilities where electronics are in scope.
23. Fixtures Are Part of the Process
Assembly fixtures locate components, support reaction loads, control geometry, protect surfaces and make the correct sequence easier. Poor fixtures transfer variability to operators.
Fixture design should consider datum strategy, tolerance stack-up, accessibility, ergonomic loading, mistake-proofing, maintenance and verification.
24. Poka-Yoke: Make the Wrong Build Difficult
Error risk | Possible mistake-proofing |
|---|---|
Wrong variant component | Barcode / keyed kit / scan interlock |
Reversed part | Asymmetric fixture / keyed interface |
Missing fastener | Count / vision / tool-cycle completion |
Wrong torque program | Automatic program selection from serial |
Unlatched connector | CPA/secondary-lock check, vision or functional test |
Missing seal | Presence sensor / vision / controlled kit |
Wrong label | Print-on-demand from serial/configuration |
POKA-YOKE PRINCIPLE
DO NOT RELY ON FINAL INSPECTION TO CATCH AN ERROR THAT THE PROCESS CAN PREVENT.
25. Traceability Architecture
Traceability should be designed around risk and customer requirements. Recording everything can create unusable data; recording too little can make containment impossible.
Traceability layer | Typical examples |
|---|---|
Product identity | Serial / batch / work order |
Configuration | BOM revision / option code / software version |
Critical components | Supplier lot / serial / date code |
Process | Station / operator / tool / program |
Critical result | Torque, force-displacement, leak, electrical test |
Material / consumable | Adhesive, sealant, grease batch where critical |
Release | Inspector / date / final status |
TRACEABILITY RULE
COLLECT DATA THAT ENABLES CONTROL, CONTAINMENT OR COMPLIANCE—not data with no decision purpose.
26. Serial Number vs Batch Traceability
Serial-level traceability is appropriate when individual configuration, safety-critical results, repair history or warranty analysis justify it. Batch-level traceability can be sufficient for lower-risk homogeneous production.
The required granularity should be agreed before system design because it changes scanning, database, labeling and operator workload.
27. Testing Should Verify Function, Not Repeat Every Upstream Inspection
Final test is strongest when it detects integration failures that cannot be efficiently controlled at component level: leakage after sealing, continuity after harness installation, actuator motion after alignment, or system response after configuration.
Do not use expensive end-of-line testing to compensate for uncontrolled upstream assembly processes.
TEST PRINCIPLE
BUILD QUALITY INTO THE PROCESS. USE TESTING TO VERIFY THE INTEGRATED FUNCTION.
28. Test Strategy: Layer the Evidence
Stage | Purpose |
|---|---|
Incoming | Verify high-risk input identity / condition |
In-process | Confirm hidden or irreversible characteristics before closure |
Process monitoring | Capture torque, force, pressure, time, etc. |
Subassembly test | Localize faults before higher-level integration |
End-of-line test | Verify integrated product function |
Final inspection | Confirm configuration, workmanship, labels, pack/release |
29. Leak Testing
Pneumatic, pressure-decay, flow, vacuum or tracer-gas methods may be appropriate depending on product and required sensitivity. The test specification should define medium, pressure/vacuum, stabilization, test time, limit, temperature considerations and fixture contribution.
A 'no visible bubbles' instruction is not equivalent to a quantified leak requirement unless that method is the approved acceptance criterion.
30. Electrical Testing
Depending on product scope, electrical verification may include continuity, shorts, insulation resistance, dielectric withstand, grounding/bonding, sensor response, power-up and functional I/O.
The exact limits and safety precautions must come from the product/test specification. Avoid inventing generic test voltages or resistance thresholds.
31. Functional Testing
Functional test should simulate the required product behavior at a level appropriate to the assembly. A motor module might be checked for rotation/current/sensor feedback; a valve assembly for actuation/leakage; a control enclosure for I/O and interlocks.
Define test coverage, fixtures, calibration, golden-reference strategy, software version, data retention and failure disposition.
32. The Golden Sample Is Useful—but Not the Specification
A boundary or master sample can clarify workmanship, routing or cosmetic expectations. It should not override released drawings/specifications or become an uncontrolled engineering baseline.
SAMPLE RULE
USE SAMPLES TO CLARIFY ACCEPTANCE—not to replace configuration-controlled requirements.
33. Final Inspection
Final inspection should focus on characteristics that remain relevant after assembly: configuration, visible workmanship, labeling, completeness, critical interfaces, test status and packaging.
Correct product/variant/serial
No missing hardware or caps/plugs
Critical witness points complete
Labels readable and correct
Test result accepted
Protective features installed
Documentation complete
Packaging configuration correct
34. Quality Gates Should Follow Irreversibility
The best inspection point is often immediately before a characteristic becomes hidden or expensive to correct. Verify an internal connector before closing the enclosure; verify seal placement before compression; verify a press fit before surrounding components block access.
GATE PRINCIPLE
INSPECT BEFORE THE DEFECT BECOMES HIDDEN.
35. Rework and Repair
Rework should return the product to the released requirement through an approved method. Repair may change the product from the original requirement and can require engineering/customer approval.
Repeated rework is also process data. A line that 'always fixes' the same issue has a process problem even if every unit eventually passes final test.
36. Nonconformance and Deviation Control
When a component or assembly does not meet requirement, disposition authority should be clear: scrap, rework, repair, return, use-as-is concession or deviation.
Temporary deviations need scope, quantity/time limit, affected configuration and expiry. Otherwise temporary changes become undocumented permanent processes.
37. Engineering Change Management
Assembly programs are particularly sensitive to change because one ECO can affect BOM, inventory, tooling, fixtures, work instructions, test software, labels and packaging.
Change question | Control |
|---|---|
When does revision become effective? | Serial / work-order cut-in |
What happens to old inventory? | Use-up / rework / scrap / return disposition |
Does fixture/tool change? | Validation before release |
Does test change? | Program/version control |
Are operators affected? | Training acknowledgment |
Are suppliers affected? | Flow-down and confirmation |
Can old/new revisions mix? | Explicit configuration rule |
38. Training and Competence
Training should match process risk. General orientation is not enough for critical torque, crimping, soldering, adhesive mixing, electrical test or inspection tasks.
Competence can be demonstrated through qualification, supervised builds, periodic verification and performance data, depending on the process and customer requirement.
39. Calibration and Measurement-System Control
Torque tools, gauges, pressure sensors, electrical testers, leak testers and other measurement devices should have defined verification/calibration status appropriate to their use.
Calibration alone does not guarantee a capable measurement system. Fixture repeatability, operator method, environmental conditions and test software can also affect results.
40. Capacity Planning
Assembly capacity is driven by takt/cycle time, station balance, labor skills, test bottlenecks, fixture count, material availability and changeover—not only floor area or headcount.
A program should identify the bottleneck operation and capacity recovery plan before volume ramp.
41. Line Balancing
If one station requires twice the time of every other station, adding operators elsewhere does not increase output. Work content should be redistributed, parallelized or automated where justified.
Testing frequently becomes the hidden bottleneck because its cycle is fixed by physics or specification.
42. Automation: Use It Where It Controls Risk or Repetition
Good automation candidate | Weak reason to automate |
|---|---|
High-volume repetitive fastening | 'Automation looks modern' |
Critical traceable torque | Product changes every few weeks |
Barcode/configuration interlock | Manual cycle is already trivial and robust |
Repeatable functional test | Test limits/design are unstable |
Vision presence/orientation | Defect has no meaningful visual signature |
Press force-displacement | No validated acceptance window |
AUTOMATION PRINCIPLE
AUTOMATE THE STABLE, REPETITIVE AND HIGH-VALUE CONTROL POINT—not unresolved engineering variability.
43. Low-Volume / High-Mix Assembly
High-mix industrial products need flexible controls: digital work instructions, scan-based variant confirmation, modular fixtures, configurable test programs and disciplined kitting.
The goal is to achieve configuration control without building dedicated automation for every variant.
44. High-Volume Assembly
At higher volume, dedicated fixtures, balanced cells, automatic tool selection, in-line verification and error-proofed material presentation can reduce recurring labor and escape risk.
Automation economics should include uptime, maintenance, engineering support and change flexibility—not only direct labor saved.
45. Packaging Is the Last Assembly Operation
Packaging protects the completed product's configuration and quality through storage, transport and line-side handling. Missing caps, poor restraint, metal-to-metal contact or wrong orientation can destroy value created by the assembly line.
Define shipping orientation.
Protect machined/cosmetic surfaces.
Cap open hydraulic/pneumatic/electrical interfaces.
Prevent movement and fretting.
Control corrosion protection where required.
Ensure labels/serials remain accessible.
Design returnable packaging when lifecycle economics justify it.
46. Documentation Pack
Depending on customer and industry, shipment or lot records can include material certificates, inspection reports, torque/test records, certificates of conformity, deviation approvals, serial lists and packaging documentation.
The required pack should be defined contractually; generating documents nobody uses adds cost, while missing required records can block acceptance.
47. ISO 9001 and Assembly Quality Systems
ISO 9001 provides a general quality-management framework for controlled processes, documented information, competence, externally provided processes/products/services, release and nonconforming outputs. As of August 2026, ISO 9001:2015 remains published with Amendment 1:2024, while ISO notes a replacement edition is under development.
ISO 9001 certification does not define the detailed assembly acceptance criteria for a specific product. The OEM still needs drawings, specifications, workmanship standards, test limits and customer-specific requirements.
QUALITY-SYSTEM GUARDRAIL
A CERTIFIED QMS CONTROLS THE SYSTEM. IT DOES NOT REPLACE PRODUCT-SPECIFIC ENGINEERING REQUIREMENTS.
48. Contract Assembly Cost Drivers
Driver | Cost mechanism |
|---|---|
BOM value / supplier count | Purchasing, working capital, incoming control |
Variant complexity | Kitting, instructions, changeover, error risk |
Assembly labor content | Direct recurring conversion |
Critical fastening / joining | Tooling, control, traceability |
Fixtures / tooling | NRE + maintenance |
Test coverage / cycle | Equipment, engineering and bottleneck time |
Traceability depth | Scanning, database, data retention |
Quality documentation | Inspection and release effort |
Volume / takt | Line design and automation economics |
Packaging / logistics | Protection, cube, racks and freight |
49. How to Reduce Assembly Cost Without Increasing Risk
Reduce part count and fastener variety through DFA.
Standardize components across variants.
Design one-direction / self-locating assembly where possible.
Move error detection upstream before enclosure/closure.
Use common torque tools/program families where feasible.
Integrate repeated tests into one controlled fixture.
Use scan-based configuration only where it prevents real risk.
Separate production batch from delivery batch when inventory permits.
Design packaging as part of the product flow.
Eliminate documentation that has no contractual/control purpose.
50. Make-or-Buy: When Contract Assembly Creates Value
Strong outsourcing case | Potential reason to retain in-house |
|---|---|
Many supplier interfaces can be consolidated | Assembly contains protected core IP/process |
OEM line space is constrained | Very low volume + constant engineering change |
Supplier has relevant process/test capability | Test depends on unavailable proprietary infrastructure |
Assembly can be fully specified and accepted | Product definition is not mature enough to transfer |
Regional supply chain can be integrated | Logistics overwhelms conversion advantage |
OEM wants delivered tested modules | Integration risk cannot yet be contractually separated |
MAKE-OR-BUY RULE
OUTSOURCE A CONTROLLED PROCESS—not an undefined problem.
51. Contract Assembly Supplier Evaluation
Can the supplier control BOM revisions and variants?
How are buyer-supplied and supplier-sourced materials segregated?
How are shortages and substitutions controlled?
Are work instructions revision-controlled at point of use?
How are torque/process programs selected and locked?
What traceability can be linked to serial number?
How are fixtures/testers validated and maintained?
How are ECOs cut into production?
How are nonconforming assemblies contained?
What is the bottleneck and backup-capacity strategy?
Can test data and quality records be exported in the required format?
How are packaging and final-release requirements controlled?
52. Worked Example A — Mechanical Subassembly
A welded frame receives bearings, shafts, guards and bolted brackets. Early builds show variable bearing preload and repeated alignment rework.
The team defines datum-based fixture location, press-fit monitoring, controlled tightening sequence and an in-process rotation check before guards hide the mechanism.
LESSON
CONTROL THE INTERFACE BEFORE IT BECOMES A SYSTEM-LEVEL FAILURE.
53. Worked Example B — Electromechanical Module
A module combines a motor, gearbox, sensor, harness and control connector. Components individually pass incoming inspection, but field failures show intermittent sensor connections.
The assembly process adds connector-lock verification, controlled routing/strain relief and an end-of-line motion + sensor-feedback test.
LESSON
COMPONENT CONFORMANCE DOES NOT GUARANTEE INTEGRATED FUNCTION.
54. Worked Example C — Wrong Variant
Two product variants differ only by one harness and one label. Operators occasionally build the wrong combination.
A scan-based kit traveler links serial number to variant, validates harness part number and prints the correct label only after test completion.
LESSON
CONFIGURATION ERRORS SHOULD BE PREVENTED—not inspected out.
55. Worked Example D — Critical Torque
A safety-relevant bracket uses several visually identical fasteners with different joint requirements. Manual click wrenches and witness paint create weak traceability.
The cell moves to controlled tooling with program selection tied to the work order, result capture and exception handling. Witness marks remain only as a visual status aid.
LESSON
THE DATA MUST PROVE THE CONTROLLED PROCESS—not merely that somebody touched the bolt.
56. Worked Example E — Leak-Test Bottleneck
Assembly labor is balanced for a 90-second takt, but the specified leak test requires substantially longer stabilization and measurement. Output misses target despite adding operators.
The solution evaluates parallel fixtures and test architecture rather than adding labor to non-bottleneck stations.
LESSON
CAPACITY IS SET BY THE BOTTLENECK, NOT BY AVERAGE WORK CONTENT.
57. Worked Example F — Engineering Change
A connector revision is approved while old inventory remains in kits, line-side bins and supplier stock. Without a controlled cut-in, mixed configurations become possible.
The ECO defines effective serial, inventory disposition, new work instruction, tester-program revision and supplier cut-in confirmation before production release.
LESSON
AN ECO IS A SUPPLY-CHAIN AND PRODUCTION EVENT—not only a drawing revision.
58. 30-Point Assembly Readiness Checklist
Released BOM and drawings exist.
Variant logic is explicit.
Approved substitutes are defined.
Supplier-sourcing responsibilities are clear.
Incoming inspection is risk-based.
Nonconforming input segregation exists.
Kitting method is defined.
Work instructions are revision-controlled.
Assembly sequence is validated.
Critical joints have engineering requirements.
Torque/press tools are controlled.
Consumables have shelf-life/lot controls.
Harness routing/connection criteria are defined.
ESD controls exist where needed.
Fixtures locate from functional datums.
Poka-yoke addresses high-risk errors.
Traceability granularity is defined.
Serial/lot architecture is agreed.
Hidden characteristics have in-process gates.
Test specification and limits are released.
Test fixtures/software are version-controlled.
Golden samples do not replace specifications.
Rework/repair authority is defined.
ECO cut-in process is defined.
Operator competence requirements are defined.
Measurement equipment is controlled.
Capacity bottleneck is identified.
Packaging is released.
Required documentation pack is defined.
FAI/pilot build precedes serial ramp.
59. What to Include in a Contract Assembly RFQ
RFQ input | Why it matters |
|---|---|
BOM + revision | Scope, procurement and configuration |
Drawings / specifications | Assembly requirements |
Annual + batch volume / variants | Line and inventory design |
Buyer-supplied vs assembler-sourced parts | Commercial responsibility |
Workmanship / industry standards | Acceptance baseline |
Critical torque / joining requirements | Tooling and traceability |
Test specification + limits | Equipment and cycle time |
Traceability requirements | MES/scanning/data architecture |
FAI / PPAP / validation | Launch scope |
Packaging / delivery format | Final process and logistics |
Quality records / retention | Documentation effort |
Forecast / ramp profile | Capacity and sourcing |
Tooling ownership | NRE and asset responsibility |
ECO / obsolescence terms | Change-risk allocation |
Frequently Asked Questions
What is contract assembly?
Contract assembly is the outsourced integration of components and subassemblies into a higher-level module or finished industrial assembly, often including sourcing, kitting, fastening, wiring, testing, traceability and packaging.
What is the difference between contract manufacturing and contract assembly?
Contract manufacturing may include making individual components through machining, casting, fabrication, molding or other processes. Contract assembly focuses on integrating those components into a controlled higher-level product; many programs combine both.
What should an OEM provide to an assembly supplier?
At minimum, a released BOM, drawings/specifications, variant logic, critical process requirements, test/acceptance criteria, volumes, sourcing responsibility, traceability and packaging requirements.
Who should own the BOM?
The design authority normally owns product definition. The assembler controls the released BOM at point of use and should not make uncontrolled substitutions.
How should torque be controlled?
Based on joint engineering requirements using appropriate tools, programs, calibration/verification, sequence and traceability. Torque alone is an indirect proxy for preload and should not be treated as a universal joint-quality measure.
Is a torque paint mark proof of correct torque?
No. It can be useful status evidence or movement indication, but it does not independently prove that the required tightening result was achieved.
What is assembly traceability?
The ability to link product identity/configuration to relevant components, lots, processes, tools, operators, test results and release records at a granularity appropriate to risk.
Should every component be serial-tracked?
No. Traceability depth should follow risk, compliance, containment and warranty needs. Recording everything can add cost without improving control.
What is kitting?
Kitting prepares the correct quantity and revision of components for a specific assembly, batch or work order before line execution, reducing search, shortages and mixed-variant risk.
What is poka-yoke?
Mistake-proofing: designing the product, fixture, tooling or information flow so an incorrect action is impossible or immediately detected.
When should an assembly be tested?
At points that efficiently verify hidden/irreversible characteristics and integrated function. Subassembly and end-of-line tests should complement process control, not compensate for uncontrolled assembly.
What tests are common in electromechanical assembly?
Depending on the product: continuity, insulation, grounding, power-up, sensor response, motion, current/voltage, functional I/O, leak/pressure and safety-interlock checks. Limits must come from the product specification.
What is IPC/WHMA-A-620?
An industry-consensus requirements and acceptance standard for cable and wire harness assemblies. The applicable revision/class/addenda and customer requirements should be specified contractually.
Does ISO 9001 define assembly workmanship?
No. ISO 9001 is a quality-management-system framework. Product-specific drawings, workmanship standards, test limits and customer requirements still define what is acceptable.
How should engineering changes be handled?
With a controlled effective point, inventory disposition, updated BOM/drawings/work instructions/test programs, supplier flow-down, operator training and configuration verification.
What is the difference between rework and repair?
Rework returns a nonconforming product to the original requirement through an approved method. Repair can make the product usable without fully restoring the original requirement and may need engineering/customer approval.
When should assembly be automated?
When the task is stable, repetitive and economically or quality-critical—such as traceable fastening, presence verification, press monitoring or repeatable testing. Automation should not freeze an unstable design.
How is assembly capacity calculated?
From takt demand, station cycle times, bottleneck equipment/testers, shift pattern, changeover, uptime, fixture count, labor competence and material availability.
How can contract assembly reduce supply-chain complexity?
One integrator can manage multiple component sources, kitting, assembly, test and final release, reducing the number of interfaces the OEM must coordinate. The benefit depends on clear responsibility and strong configuration control.