Product
- Complete robot platform, ODM design, private-label product or configured subsystem.
Robotics Supply-Chain Risk Guide
How to map China-based components, factories, materials and sub-tier dependencies—and evaluate China Plus One without creating false diversification
China dependency cannot be measured from direct-supplier nationality or final-assembly location alone.
A robot may be assembled outside China while still depending on China-based reducers, motors, electronics, tooling, magnets, calibration equipment, sub-tier suppliers or manufacturing know-how.
This guide maps dependency through the BOM, production sites, processes, materials, tooling, software, calibration, testing and replacement lead time.
The practical objective is not to label a country as good or bad. It is to know which exact dependencies can stop production, delay field service, force redesign, change compliance evidence or reduce product performance, and which controls are strong enough to keep the robot program operating under stress.
China supply-chain dependency in robotics is the degree to which a robot product, component, process, material, production site, tool, software configuration or lifecycle service relies on China-based capability that cannot be replaced quickly without unacceptable cost, performance loss, quality risk or redesign.
The dependency may sit at final assembly, but it often sits upstream: a reducer factory, servo-motor process, encoder supplier, PCB assembly line, battery-cell source, magnet supplier, precision machining cluster, calibration fixture, test software, outsourced heat-treatment shop or one team of engineers that knows how to tune the production process.
Dependency is therefore not determined by the nationality of the direct supplier. It is determined by the controlled path from the robot BOM to manufacturers, production sites, sub-tier suppliers, processes, materials, tooling, software, calibration, testing and replacement lead time.
This is an engineering and sourcing framework, not a political article, country ranking, supplier directory, legal opinion or warehouse-robotics guide. It applies to the supply chain used to manufacture robots and robotics components. For the broader hub, start with robotics supply-chain mapping.
A strong dependency assessment should leave the team with a working map, a risk register, and a prioritized action plan. It should be specific enough that engineering can review substitution impact, sourcing can negotiate disclosure and continuity terms, quality can define approval evidence, and operations can plan inventory, transfer or recovery work without relying on broad assumptions.
The terms are related but not interchangeable. A robotics company can be exposed to China without being dependent on China for a specific component, and it can have concentrated supply even when the country is not the central issue.
Keeping the terms separate avoids two opposite errors. One error is overreacting to any China touchpoint, even when the item is noncritical and readily replaceable. The other is underreacting because there are multiple supplier names, while all of them rely on the same factory, special process, industrial cluster, software tool or material source.
| Term | Meaning in robotics sourcing | Example | Common mistake |
|---|---|---|---|
| Exposure | A product, supplier, material or route touches China-based activity. | A camera module is purchased from a distributor that sources from a China-based assembly site. | Treating any exposure as severe dependency. |
| Concentration | Supply is clustered in one country, region, facility, process, supplier group or material source. | Three quoted suppliers for a machined joint housing all outsource anodizing to the same industrial cluster. | Counting supplier names instead of production sites and sub-tiers. |
| Dependency | Replacement is not available within acceptable time, cost, performance and qualification limits. | A calibrated actuator module uses a China-based reducer supplier with no approved alternative. | Assuming dependency disappears when final assembly moves. |
| Risk | The likelihood and consequence of a dependency becoming a business problem under defined scenarios. | A port closure, export restriction, supplier shutdown or component redesign delay threatens production continuity. | Assigning risk without defining scenario, impact and controls. |
describes risk management as a structured process of identifying, analyzing, evaluating and treating risk. In this guide, the same discipline is applied to product-specific China dependency rather than to a generalized country label.
China matters because it combines large robotics demand, broad component ecosystems, mature electronics and battery supply chains, dense mechanical-processing clusters, export-oriented factories and a large base of suppliers that can build robot subsystems or complete products.
The International Federation of Robotics reported that China accounted for 54% of global industrial-robot installations in 2024, with 295,000 units installed. IFR also reported that Chinese manufacturers reached 57% domestic market share in China. These figures show the scale and maturity of China’s robotics market and domestic supplier base.
Those market figures do not measure product-specific dependency. A company cannot infer its own exposure from national installation share, and it should not assume that a China-made robot or component is fully China-independent. A supplier in China may depend on imported bearings, encoders, processors, optical components, software tools or metrology equipment. Conversely, a supplier headquartered outside China may manufacture a critical component in China or depend on China-based sub-tiers.
For sourcing teams, the importance of China is operational. The country may offer dense supplier discovery, fast prototyping support, mature electronics ecosystems, battery and cable supply, precision mechanical processing, and suppliers willing to integrate subsystems. Those strengths can be valuable. They also make it easy for dependencies to accumulate quietly across many small decisions unless the robot architecture is mapped deliberately.
Use market data as context, not as a dependency score. Dependency must be mapped from the robot BOM, the production sites and the sub-tier network used for the specific product.
A robotics product can depend on China at many layers. The buyer should define which layer is being evaluated before discussing China Plus One, second sourcing, inventory or transfer.
The same robot may have different dependency states by layer. A complete robot platform may be diversified at final assembly, concentrated for actuators, exposed for battery cells, dependent for calibration software, and transfer-ready for machined covers. Treating these as one blended country question hides the actions that would actually reduce risk.
China dependency mapping begins with the product and moves upstream. The goal is a visible, crawlable dependency structure that separates manufacturer, supplier, site, process, material, tooling, software and recovery options.
The map should be built to support decisions, not merely documentation. Each record should answer what would happen if the China-based path were unavailable, how quickly the team would know, who owns the response, what evidence is needed to switch source, and which customer, quality or regulatory approvals would be affected by the change.
A dependency map should be stored as structured records, not as a narrative memo. The record should allow a team to trace one robot configuration from a BOM item to supplier roles, production sites, sub-tier inputs, risk state and mitigation status.
Structured records are especially important when a robot has several variants, customer-specific configurations or frequent engineering changes. Without a controlled data model, one team may analyze an old BOM, another may qualify a different supplier revision, and operations may buy material against an assumption that no longer matches the released product.
| Field group | Record fields |
|---|---|
| Product baseline | Robot model, variant, revision, destination market, production phase, owner, review date. |
| BOM item | Part number, description, function, subsystem, criticality class, quantity, revision and approved manufacturer. |
| Supplier role | Contracted supplier, actual manufacturer, distributor or trading company status, legal entity and headquarters. |
| Manufacturing site | Country, region, city, site address, site owner, internal process, outsourced process and capacity constraint. |
| Sub-tier | Tier level, supplier name, supplied item, production site, single-source status and visibility confidence. |
| Process | Critical process, equipment, tooling, fixture, calibration method, test method and process owner. |
| Material | Critical material, material source, refining or conversion source, specification and substitution constraint. |
| Software | Firmware version, driver, library, license, configuration file, update owner and support status. |
| Resilience | Approved alternatives, qualification status, replacement lead time, inventory coverage, transfer package and recovery scenario. |
| Governance | Dependency state, risk category, mitigation owner, status, evidence source, confidence level and next review date. |
Visible fields also reduce ambiguity in supplier discussions. If a supplier says a component is “local,” the map should identify whether that means local sales office, local final assembly, local manufacturer, local raw material, local process capability or local lifecycle support.
The BOM is the only defensible starting point because it describes what the robot actually uses. A supplier list starts from commercial relationships; a BOM starts from the product architecture and exposes what must be available for production and service.
A BOM-first method also prevents the team from focusing only on the most visible suppliers. The dependency that controls recovery may be hidden inside a joint module, a battery pack, a controller board or a preassembled cable harness. By expanding each critical BOM item into manufacturer, site, process and sub-tier records, the map follows the product rather than the purchasing hierarchy.
Robots are systems. A small part by cost can be decisive for safety, motion accuracy, calibration stability, firmware compatibility or field repair. A low-value connector, encoder, seal, magnet, bearing, PCB component or calibration target can stop production if it is the only approved item and the replacement requires engineering validation.
BOM value is not the same as dependency. A percentage of spend from China can be useful context, but it cannot replace component-level analysis of criticality, substitutability, process control and recovery time.
The BOM should connect to component guides such as robot components, robot actuators, harmonic reducers, servo motors, robot controllers, robot sensors and machine vision.
Sub-tier suppliers matter because the direct supplier may not control the technology that creates dependency. A Tier-1 actuator module supplier can depend on a separate reducer factory, motor winding line, encoder supplier, bearing source, drive PCB assembler, magnet supply chain and calibration fixture.
Sub-tier visibility does not mean asking every supplier to disclose every commodity screw. It means identifying the lower-tier inputs that control performance, continuity, compliance or replacement difficulty. The depth should be risk-based: a safety sensor, harmonic reducer or controller processor deserves more visibility than a standard label or generic packaging item.
| Distinction | Why it matters |
|---|---|
| Manufacturer vs distributor | A distributor can provide supply access, but it may not control production, change notification or recovery. |
| Headquarters vs manufacturing location | A non-China headquarters can still rely on a China-based production site or sub-tier. |
| Final assembly vs upstream independence | Moving final assembly may leave key components, tooling and testing in China. |
| Candidate vs qualified source | A market candidate does not protect production until it is approved for the product and process. |
Motion systems are often the highest-consequence dependency in robotics because they combine mechanical precision, electromechanical performance, control tuning, calibration, reliability and capacity. The relevant unit of analysis is usually the joint architecture, not a single purchased line item.
Motion dependency should be evaluated with engineering, not purchasing alone. A replacement reducer may fit the envelope but change stiffness, backlash or life. A replacement motor may meet nominal torque but alter thermal behavior or control tuning. A replacement encoder may require firmware and calibration changes. The supply-chain decision is therefore also a product-performance decision.
| Motion item | China dependency questions | Replacement impact |
|---|---|---|
| Actuator module | Where is final module assembly, calibration and end-of-line testing performed? | Mechanical interfaces, firmware, torque control, thermal model and validation. |
| Reducer | Who makes the flexspline, wave generator, gears or precision transmission components? | Backlash, stiffness, life, noise, lubrication, mounting and qualification time. |
| Servo motor | Where are winding, magnet insertion, rotor balancing and test performed? | Torque constant, thermal behavior, control tuning and safety margin. |
| Encoder | Where are sensing elements, optics, magnetic scales or calibration produced? | Resolution, accuracy, firmware drivers, zeroing and control stability. |
| Drive electronics | Where are PCBs assembled, firmware loaded and power components sourced? | EMC, thermal design, firmware, safety functions and component lifecycle. |
For component-level detail, see robot actuators, harmonic reducers and servo motors.
Controller dependency can sit in board design, semiconductor sourcing, PCB fabrication, SMT assembly, firmware provisioning, safety logic, motion-control libraries, cybersecurity maintenance or the production test system used to release the controller.
A controller assembled outside China may still depend on China-based PCB fabrication, connectors, cable assemblies, power supplies, housings or subassembly test fixtures. A controller sold by a China-headquartered supplier may also depend on imported processors, memory, power semiconductors, operating systems or development tools.
Electronics dependency also changes over time because component lifecycles are shorter than mechanical lifecycles. A robot platform may remain mechanically stable for years while processors, memory, connectors, wireless modules or power devices approach end of life. The China map should therefore connect electronics dependency to obsolescence monitoring and redesign lead time.
Technical selection depth belongs to the robot controllers guide.
Sensors and vision systems create dependency through precision elements, calibration, software drivers, optics, illumination, embedded processors and field-reliability behavior. They are frequently purchased as catalog parts, but catalog availability does not mean substitution is low risk.
For perception-heavy robots, the data output can matter as much as the physical sensor. Replacing a camera, LiDAR unit or force sensor may shift noise, latency, calibration drift, mounting alignment or driver behavior. The supply-chain map should therefore record not only the source of the sensor but also the software, calibration and validation evidence tied to that source.
Encoders, force-torque sensors, IMUs and tactile sensors can depend on specialized calibration and stable supplier revisions.
Replacement may affect safety evidence, sensing range, environmental robustness and regulatory documentation.
Image sensors, lenses, filters, lighting and camera firmware can have separate countries of manufacture and lifecycle exposure.
Drivers, SDKs, AI models and accelerator hardware can become the controlling dependency even when the camera body is available.
See robot sensors and machine vision for component selection and validation considerations.
Mobile robots, service robots, humanoids and portable platforms can depend on China for cells, battery-pack assembly, BMS electronics, chargers, connectors, thermal materials, safety testing, transport documentation and recycling or service pathways.
The dependency should separate cell origin from pack assembly and from BMS ownership. A pack assembled in a second country may still use China-origin cells, China-made protection electronics or China-based pack tooling. Conversely, a China-based pack supplier may use imported cells or foreign BMS chips.
Battery decisions also affect logistics and lifecycle support. A change in cell, pack design, charger or BMS firmware can affect transport classification, storage procedures, safety documentation, field-service instructions and warranty policy. A second source for batteries must therefore be validated as a complete power-system and service decision, not merely as an alternate pack quotation.
| Power element | Mapping question | Risk if unmapped |
|---|---|---|
| Cells | Who manufactures the cells and where are they produced? | Qualification and transport assumptions may be wrong. |
| Pack assembly | Which site welds, assembles, tests and labels the pack? | Final-pack relocation may fail due to tooling or process gaps. |
| BMS | Who owns the board design, firmware and configuration? | Replacement may require software, safety and charging changes. |
| Charger and power supply | Where are power electronics and compliance-critical parts sourced? | Regulatory evidence and field compatibility can be affected. |
Mechanical and interconnect items are sometimes treated as simple commodities, but they often contain hidden dependency. Robot frames, housings, castings, precision-machined parts, cable harnesses, flex circuits, connectors, slip rings and seals affect reliability, safety, serviceability and production flow.
These parts also tend to expose the difference between a drawing and a production process. A frame may require a specific casting source, stress-relief method, machining sequence or inspection fixture. A harness may require approved crimp tools, bend-radius controls and flex-life materials. If those production details remain undocumented, the part is harder to move than its apparent simplicity suggests.
Critical-material dependency should be mapped carefully and without overclaiming. Not every robot motor uses rare-earth permanent magnets, and not every China dependency in motion systems is a magnet dependency. Motor architecture, torque density, cost target, duty cycle and supplier design choices determine whether rare-earth magnet exposure is material.
The International Energy Agency has highlighted the concentration of rare-earth supply chains used in permanent magnets, including roughly 60% of mined magnet rare earths, more than 90% of refining, and about 95% of permanent-magnet production associated with China. These figures are important context for motors, generators and other magnet-dependent products, but they do not prove that every robot or every motor has the same exposure.
Material mapping should connect the material to the component and performance requirement. If a servo motor uses a specified magnet grade, the map should identify whether an alternate grade, supplier or motor design is already validated. If no alternate exists, the dependency may be a design constraint rather than a purchasing constraint, and mitigation may require redesign or supplier development.
Map the actual material path. Identify whether the robot uses permanent-magnet motors, which magnet grade is specified, who makes the magnet, where raw materials are mined and refined where relevant, and whether the motor design has qualified alternatives.
provides guidance for integrating sustainability into procurement decisions. Responsible sourcing should be connected to actual materials and suppliers rather than to broad assumptions about a product category.
Tooling and equipment can be the reason a second source is not usable. A buyer may own the design but not own, control or be able to move the tooling, fixtures, gauges, machine programs or calibration stations needed to manufacture the product.
The assessment should separate ownership, physical location, technical capability and practical transferability. A contract may say the buyer owns tooling, but the tool may be embedded in a supplier process, worn beyond useful transfer, dependent on local maintenance knowledge, or useful only with a machine that the alternative site does not have.
| Asset | Dependency question | Transfer implication |
|---|---|---|
| Production tooling | Who owns the mould, die, fixture, winding tool or assembly jig? | Transfer may require legal rights, refurbishment, duplicate tooling or redesign. |
| Process equipment | Is the capability common, specialized or supplier-developed? | A new site may lack equipment accuracy, operator skill or maintenance support. |
| Inspection gauges | Are gauges controlled, calibrated and transferable? | Quality approval may not be comparable across sites. |
| Calibration station | Who built the station and owns the software? | Performance release may remain dependent on the original supplier. |
| Test bench | Does end-of-line testing rely on supplier-specific hardware or scripts? | A second source may build parts but not release conforming products. |
Yes. Manufacturing knowledge can be a dependency when production performance depends on undocumented supplier know-how, process tuning, operator skill, engineering judgment or calibration interpretation located in China.
This is especially relevant for precision reducers, compact actuators, high-density motor assemblies, sensor calibration, battery-pack process control, cable routing, thermal management and end-of-line testing. A drawing package may not contain the tacit knowledge needed to achieve stable yield, noise, backlash, torque ripple, thermal margin or field reliability.
Robotics dependency is not only physical. Firmware, drivers, calibration files, production software, model parameters, programming fixtures, licenses, certificates and test scripts can bind a robot to a supplier or location.
This layer is often discovered late because it is not always represented in the purchasing BOM. A supplier may ship a working actuator or controller, but the ability to program, tune, calibrate, sign firmware, interpret test data or issue updates may remain with the original factory. A China Plus One site without this access may assemble hardware but still depend on China for release.
A relocation or China Plus One program should ask whether the new site can provision firmware, perform calibration, run the same test method, interpret failures, store records and support field updates. If not, the product may have moved physically while remaining operationally dependent on the original supplier.
Logistics dependency covers transportation routes, ports, freight forwarders, customs processes, warehousing, export documents, packaging and temperature or hazardous-goods constraints. Robotics products may also have large, delicate or regulated components that make route changes harder than standard electronics shipments.
Inventory is a resilience control, not a dependency cure. Buffer stock can protect production during a short disruption, create time for qualification, or support a planned transfer. It does not create another qualified supplier, move tooling, change material origin, resolve firmware ownership or validate an alternate process.
Inventory creates time. It does not eliminate dependency unless the underlying supply path is replaced or a qualified second source is released.
can inform business-continuity planning, while includes supply-chain-relevant security-management aspects. Applicability depends on the company’s product, customers, contracts and risk profile.
Yes. China dependency mapping should be bidirectional. A China-based robotics supplier may depend on foreign semiconductors, bearings, encoders, sensors, optical components, industrial software, metrology equipment, machine tools, materials or customer-provided intellectual property.
Bidirectional dependency matters for continuity planning because the failure mode may not originate in China. A foreign processor allocation, machine-tool spare part, software-license issue or customer-controlled design change can disrupt a China-based supplier. The map should therefore identify both what the buyer depends on in China and what the China-based supplier depends on outside China.
This matters because a product can be dependent in both directions. The buyer may depend on a China-based supplier for production, while the supplier depends on foreign inputs for critical performance. A disruption to either side can affect robot availability.
Supplier identity must be resolved at entity level. A website, brand name, export company or sales office may not be the entity that owns the factory, controls the process, holds the tooling, signs the contract, invoices the buyer or makes engineering-change decisions.
This distinction becomes critical when a problem occurs. The entity that answers sales emails may not be able to approve a process change, release tooling, open sub-tier records, authorize an audit, or guarantee capacity. A useful map should show who can make each decision and what evidence supports that conclusion.
Public identity used in sales material; may not be the contracting entity.
Company that signs the purchase agreement and carries legal obligations.
Entity handling export, invoicing, customs and payment; may differ from factory owner.
Physical site performing manufacturing, calibration, testing or final assembly.
Affiliates that own equipment, tooling, intellectual property or sub-processes.
External supplier that controls an essential component, material or process.
Entity resolution belongs with supplier qualification and factory assessment. See robotics supplier qualification and factory capability assessment. For supplier landscape context, see robotics suppliers and China robotics manufacturers.
China should not be treated as one homogeneous manufacturing location. Robotics supply chains often concentrate by region, industrial cluster, process ecosystem and logistics route. Several supplier names may still depend on the same machining cluster, electronics assembly base, battery ecosystem, casting supplier, port or pool of specialized engineers.
Regional mapping should be practical rather than encyclopedic. The team needs enough detail to know whether a disruption, capacity constraint or supplier change would affect one factory, one city, a broader process cluster or several nominally independent suppliers. That level of detail is what turns geographic information into an operational continuity plan.
Cluster mapping prevents false diversification. Two suppliers may be commercially independent but operationally exposed to the same regional disruption, material supplier or process bottleneck.
Do not publish an unsupported single China-dependency percentage for a whole company. A single percentage is usually misleading because it blends BOM value, supplier count, revenue, spend, geography and criticality without explaining whether a replacement is qualified.
Measurement should support prioritization. A company may decide that an exposed but noncritical consumable only needs monitoring, while a small China-dependent encoder requires second-source qualification. The report should make those differences visible so leadership can fund targeted actions instead of reacting to one aggregate number.
| Dimension | What to measure | Why it matters |
|---|---|---|
| BOM exposure | Critical and noncritical items with China-based manufacturer, site or sub-tier. | Shows where product architecture touches China. |
| Criticality | Safety, performance, quality, supply, cost and lifecycle impact. | Separates low-consequence exposure from production-critical dependency. |
| Concentration | Number of qualified manufacturers, sites, regions and sub-tiers. | Identifies hidden single-source and cluster dependency. |
| Substitutability | Approved alternatives, redesign need, interface compatibility and validation burden. | Determines whether replacement is realistic. |
| Recovery time | Time to use inventory, qualify alternative, transfer tooling or redesign. | Connects dependency to business continuity. |
| Control rights | Tooling ownership, documentation, software access, change rights and supplier contracts. | Shows whether the buyer can act during disruption. |
| Evidence confidence | Supplier claims, documents, observations, records and independent verification. | Prevents false precision from weak data. |
A dashboard can summarize these dimensions, but the underlying record must remain component-level and evidence-based.
The Yana dependency-state model describes maturity of understanding and mitigation. It is not a country score and should not be collapsed into a single percentage without context.
The team does not know whether China-based suppliers, sites, materials, tooling or sub-tiers are involved.
China-based activity exists, but dependency severity and replacement readiness are not yet proven.
Multiple critical items or processes are concentrated in China-based suppliers, regions or clusters.
Production, performance or lifecycle support cannot be maintained within acceptable limits without China-based capability.
Controls such as inventory, monitoring, contracts, documentation or contingency plans reduce near-term consequence.
Additional qualified suppliers, sites or materials exist, with validated quality and capacity.
Documentation, tooling, software, process controls and qualification evidence support a controlled move to another site.
Useful metrics should help teams make decisions without pretending that dependency is a simple spend percentage. A low-cost item may be irreplaceable; a high-cost item may have qualified alternatives.
Metrics should be refreshed when the product changes. A redesign, supplier engineering change, new customer market, volume ramp, tariff exposure, component end-of-life notice or sub-tier substitution can change the state of dependency even if the direct supplier remains the same. Static dashboards create confidence only when they are tied to update triggers.
BOM value does not equal dependency. Use value-weighted views only alongside criticality, substitutability, qualification state and recovery time.
A China Plus One strategy is a diversification approach in which a company maintains some China-based supply or manufacturing capability while developing additional qualified capability outside China. This aligns with the way UNCTAD and trade-development discussions commonly describe China Plus One: diversification of production or sourcing beyond China, not necessarily exit from China.
For robotics, China Plus One should create an additional qualified production path for defined products, components or processes. It succeeds only when the alternative can meet technical requirements, quality controls, calibration, testing, capacity, lifecycle support and commercial conditions.
The strategy should also state what will remain in China by design. Some companies may intentionally retain China-based suppliers for cost, capability, speed or market access while reducing single points of failure. That can be a valid resilience design if the remaining dependencies are known, monitored and supported by contingency plans.
Not necessarily. China Plus One can reduce concentration and improve resilience, but it does not automatically remove dependency. Many programs move the last assembly step while critical components, materials, tooling, firmware or test processes remain in China.
| Moved outside China | What may remain China-dependent | How to verify |
|---|---|---|
| Final assembly | Actuators, PCBs, batteries, sensors, cables, fasteners, fixtures and end-of-line test equipment. | Map BOM manufacturers and production sites, not only assembly location. |
| Supplier contract | Same supplier group uses China-based sub-tiers or ships semi-finished modules. | Resolve legal entity, factory and sub-tier sources. |
| Machining | Heat treatment, coating, casting, tooling or metrology remains in China. | Map process flow and outsourced special processes. |
| Battery pack | Cells, BMS electronics, nickel tabs, test fixtures or transport documentation. | Separate cell origin, pack assembly and BMS ownership. |
| Controller box-build | PCBs, semiconductors, firmware programming, cable assemblies or power supplies. | Review electronics supply chain and firmware provisioning. |
| Alternative supplier | Market candidate lacks qualification, capacity or lifecycle support. | Complete sample, process, quality and capacity approval. |
Implementation should be managed as a production-readiness and supplier-qualification program. The decision is not “which country” in the abstract; it is which qualified production path can make the defined robot or component with acceptable risk.
The program should have gates. Discovery should not be called qualification, qualification should not be called production release, and pilot success should not be treated as available capacity. Each gate should produce evidence that engineering, quality, operations and sourcing can review before the next commitment is made.
Compare alternatives by product-fit criteria, not by country rankings. A country with an attractive headline may still lack the specific supplier, process, tooling, calibration, capacity or sub-tier base required for a robotics product.
The comparison should also include the cost of learning. A new manufacturing ecosystem may need supplier development, process documentation, fixture duplication, operator training, engineering support and repeated pilot builds before it reaches stable production. Those investments can be justified, but they should be visible in the decision instead of hidden behind a unit-price quote.
| Criterion | Comparison question |
|---|---|
| Technical capability | Can the supplier make the required component or robot, not just a related product? |
| Process maturity | Are the critical processes, special processes and test methods controlled? |
| Quality system | Are change control, traceability, corrective action and supplier management effective? |
| Sub-tier availability | Are local or regional suppliers available for critical inputs, or are China inputs still required? |
| Tooling and transfer | Can tooling, fixtures, gauges and production software be moved or duplicated? |
| Capacity and lead time | Is usable capacity available at the required yield and ramp rate? |
| Total landed cost | What is the combined effect of price, freight, tariffs, inventory, quality and engineering support? |
| Regulatory and compliance | Are trade, export-control, product, battery, cybersecurity or documentation issues understood? |
| Lifecycle support | Can the alternative support spare parts, repair, updates and obsolescence management? |
See alternative robotics suppliers for the full second-source evaluation framework. Cost depth belongs to robotics manufacturing cost analysis. Regulatory topics should be reviewed at a high level with robotics regulatory risks; this guide is not legal advice.
De-risking should be targeted. OECD-style discussions of supply-chain resilience generally favor targeted de-risking and diversification for critical dependencies rather than blanket reshoring or broad decoupling claims. For robotics, the practical question is which component, process, material, site or knowledge dependency creates unacceptable consequence.
Targeted de-risking also protects engineering focus. Trying to replace every China touchpoint at once can consume resources without improving the most important risks. A better sequence is to identify the dependencies that control production continuity or product performance, mitigate those first, and leave low-consequence exposures under monitoring.
Identify China-based suppliers, sites, sub-tiers, materials, tooling and software dependencies.
Track lead times, lifecycle notices, supplier changes, capacity and regulatory signals.
Use inventory, safety stock, framework agreements or reserved capacity for critical items.
Approve alternative components, suppliers, sites or sub-tier sources through evidence-based qualification.
Change interfaces, specifications or architecture to use more available parts and processes.
Build a controlled package for moving production, tooling, software, calibration and testing to another qualified site.
An alternative becomes a usable second source only when it is technically, commercially and operationally ready to produce the defined item under controlled conditions. A website, quotation, sample, distributor listing or prototype is not enough.
The distinction between market candidate and qualified second source is central to robotics supplier qualification and robotics supplier sourcing.
A manufacturing transfer requires more than a purchase order to a new factory. The package must let another site reproduce the product, process, configuration, calibration, test result and quality evidence.
The transfer plan should also define what will be revalidated. Some changes may require only process confirmation, while others affect safety, performance, reliability, regulatory files or customer approvals. Treating all transfers as simple address changes creates avoidable launch and field risks.
Production transition from prototype to controlled manufacturing is covered in robot prototype to production. Quality-system depth belongs to robotics quality systems.
Scenario tests turn the dependency map into a resilience plan. Each scenario should identify affected BOM items, time to impact, available inventory, decision owner, recovery action and validation requirements.
Regulatory scenario work should be reviewed with qualified legal or trade-compliance support where necessary. This guide provides sourcing structure, not legal advice.
| Failure | Consequence | Better approach |
|---|---|---|
| Using supplier nationality as the dependency metric | Manufacturing sites and sub-tiers remain hidden. | Map manufacturer, site, sub-tier and process. |
| Counting final assembly as independence | Critical components remain China-dependent. | Start from the BOM and process flow. |
| Confusing distributor with manufacturer | Control and change authority are misunderstood. | Resolve supplier role and entity ownership. |
| Counting market candidates as second sources | Recovery plans fail when alternatives are not qualified. | Require product, process, quality and capacity approval. |
| Ignoring sub-tier concentration | Multiple Tier-1 suppliers share one upstream dependency. | Map critical Tier-2 and Tier-3 sources. |
| Using BOM value as the main metric | Low-cost critical items are missed. | Combine value with criticality and recovery time. |
| Assuming inventory eliminates dependency | Risk returns when stock is consumed. | Use inventory as time for qualification or transfer. |
| Leaving software out of the map | Firmware, calibration and test dependencies persist. | Include software, configuration and update authority. |
| Ranking countries instead of evaluating capability | Selection becomes generic and unsupported. | Compare specific suppliers and sites against product requirements. |
| Publishing a single unsupported dependency percentage | Decision makers get false precision. | Report dimensions, evidence confidence and dependency state. |
A risk register should connect dependency evidence to ownership, status and action. It should be reviewable by engineering, sourcing, operations, quality and leadership.
The register should be used in sourcing and engineering reviews, not stored as a one-time report. When a new supplier is proposed, a BOM revision is released, or a China Plus One pilot is started, the relevant register entries should be updated so decisions are made against current dependency evidence.
| Column | Recommended values or notes |
|---|---|
| Risk ID and title | Unique identifier linked to product, subsystem and BOM item. |
| Category | Component, material, process, site, tooling, software, calibration, logistics, regulatory, lifecycle or knowledge. |
| Dependency state | Unmapped, China-exposed, China-concentrated, China-dependent, mitigated, diversified or transfer-ready. |
| Evidence source | Supplier disclosure, contract, drawing, audit record, factory observation, test record or independent verification. |
| Impact | Production stop, redesign, quality risk, cost increase, service failure, compliance review or customer delivery risk. |
| Recovery time | Inventory days, alternative qualification time, tooling duplication time and redesign time. |
| Mitigation | Inventory, second source, redesign, contract control, supplier development, transfer package or monitoring. |
| Status fields | Open, investigating, mitigation planned, in qualification, accepted, transferred, closed or monitoring. |
| Owner and review date | Named owner, function, next review date and escalation trigger. |
Supplier data requests should be specific enough to resolve entity, site, component, process and sub-tier control without asking for irrelevant information. The request should also state how confidential information will be protected.
It is reliance on China-based components, suppliers, production sites, materials, processes, tooling, software, calibration, testing or lifecycle support that cannot be replaced within acceptable time, cost and risk limits.
No. Direct-supplier nationality may hide manufacturing location, sub-tier suppliers, imported inputs, tooling ownership and process control.
No. Final assembly can move while actuators, controllers, sensors, batteries, cables, materials, tooling, firmware or test systems remain China-dependent.
The BOM shows what the robot actually uses. Supplier lists and spend summaries can miss low-cost but critical parts that stop production or require redesign.
Exposure means China-based activity is present. Dependency means the activity cannot be replaced within acceptable limits for time, cost, performance, quality and qualification.
China Plus One is a diversification strategy that keeps some China capability while developing additional qualified sourcing or production capability outside China.
Not automatically. It may leave China-based components, materials, sub-tiers, tooling, software or calibration in place. The remaining dependencies must be mapped.
No. A candidate becomes a usable second source only after technical, process, quality, capacity, commercial and lifecycle requirements are approved.
Yes. China-based robotics suppliers may depend on imported semiconductors, bearings, sensors, optical parts, software, machine tools or customer designs.
No. Some robots and motors use rare-earth permanent magnets, but the actual motor architecture and material specification must be checked before assigning magnet dependency.
Sub-tiers may control reducers, motors, cells, PCBs, sensors, materials or special processes even when the Tier-1 supplier appears diversified.
Multiple suppliers can share one industrial cluster, outsourced process, logistics route or local labor pool. Cluster mapping prevents false diversification.
No. Inventory creates time to react. It does not create qualified alternatives, move tooling, replace materials or transfer software and calibration capability.
Request legal entity, production site, manufacturer role, sub-tier sources, critical materials, process flow, tooling ownership, software, calibration, testing, capacity and continuity evidence.
Compare specific suppliers and sites against technical capability, process maturity, quality, sub-tier availability, capacity, cost, regulatory exposure and lifecycle support. Do not rank countries generically.
Transfer readiness means product documentation, process controls, tooling, software, calibration, testing, quality evidence and supplier packages are sufficient to move production under control.
Not unless the methodology is transparent and limited. A single unsupported percentage usually hides criticality, substitutability, sub-tier risk and recovery time.
No. It is an engineering sourcing framework. Regulatory and trade issues should be reviewed with qualified advisers where necessary.
Review it whenever the BOM, supplier, site, volume, market, regulation, component lifecycle or risk scenario changes, and at a defined periodic cadence for critical items.
Start by freezing the product BOM, classifying critical items, identifying actual manufacturers and sites, then mapping sub-tiers, processes, tooling, software and qualified alternatives.
Yana can support component-level dependency mapping, China Plus One evaluation, second-source development and supplier qualification for robotics products and subsystems.