EN
RUA localisation score can count operations and rights, but a vehicle experiences only behaviour. An electronic control unit must start, communicate, diagnose and remain safe through heat, vibration, voltage disturbance, software change and component substitution. Russia's proposed automotive-electronics points system therefore mattered most as an evidence ladder connecting legal control, engineering work, manufacturing and validation.
The proposal was a draft, not a completed rule
On 18 August 2026, Interfax reported that Russia's Ministry of Industry and Trade had prepared a draft points system for localising key automotive electronics. The proposal was published for public discussion through 2 September and would supplement Government Resolution No. 719.
The reported scope included navigation equipment, GLONASS and GPS modules, electronic sensors, alcohol interlocks and the software needed by those products. These categories span different safety consequences, environmental loads, production volumes and update cycles, so a common score cannot replace product-specific qualification.
Because the measure was still a draft, its reported thresholds and dates described a proposed compliance path rather than a final entitlement. Producers and customers would need to track the adopted text, transitional provisions, verification procedure and any category definitions before committing commercial assumptions.
That legal caution does not reduce the industrial value of the architecture. The draft placed rights, design, board production, software and testing on one ladder. It shifted the localisation question away from the final assembly address towards the capabilities that let a product be understood, changed and sustained.
A minimum score included mandatory foundations
Interfax said recognition would require at least 140 points together with mandatory ownership of rights to design documentation and software. The draft also required confirmation that circuit design, printed-circuit-board topology and embedded-software design were localised, and that surface mounting of chip components occurred in Russia.
This is not simply a menu where any activity can offset every missing capability. A mandatory floor defines non-substitutable foundations; additional operations build above it. That distinction prevents a large amount of low-complexity handling from compensating for absent design control.
The exact evidence accepted by the final mechanism will matter. A statement of ownership, a source-code repository, signed design records, manufacturing travellers and traceable production lots answer different questions. An audit package should connect them rather than present unrelated documents.
An evidence ladder for one electronic module
- Identify the product configuration, intended vehicle functions and accountable rights holder.
- Trace requirements into circuit, board, enclosure, interfaces and embedded-software baselines.
- Show controlled local engineering changes with review, testing and release authority.
- Link surface-mount and assembly records to component lots and process parameters.
- Demonstrate functional, environmental and vehicle-level results for the released version.
- Maintain the evidence through updates, substitutions, field incidents and obsolescence.
Rights mattered because products continued after launch
Ownership or sufficient control of documentation and software determines whether a producer can correct a defect, adapt a component, support a vehicle programme or move manufacturing. A factory that can assemble but cannot lawfully inspect or change the design remains dependent at the most consequential moments.
Rights need usable scope. Engineering must know which source files, models, libraries, test assets, build tools and third-party dependencies are included. A broad contractual phrase may still leave a critical compiler, security key or proprietary component model outside practical control.
Version completeness is equally important. A repository without the released build instructions and calibrated parameters may not reproduce the field product. The evidence baseline should include source, configuration, tool versions, binary identity, signing chain, manufacturing data and approved deviations.
The draft reportedly broadened eligible rights holders beyond legal entities to qualifying individual entrepreneurs and Russian-resident citizens. That could recognise different development structures, but product governance still needs durable transfer, access, succession and support arrangements that survive a person's absence.
A schematic and a board topology captured different knowledge
The electrical schematic describes functional connections and component intent. Printed-circuit-board topology determines physical placement, routing, layer stack, return paths, thermal behaviour and electromagnetic performance. Possessing one without the other leaves a major gap between concept and manufacturable hardware.
Automotive electronics operate amid voltage transients, switching noise, temperature cycles, moisture, vibration and constrained packaging. Layout choices influence signal integrity, heat, immunity and durability. Local topology capability therefore matters not as drawing labour but as control of physical behaviour.
Design evidence should show requirements, simulations where used, peer review, design-rule checks, prototype findings and approval. A file created locally is weak evidence if critical choices were copied without understanding or cannot be changed safely.
Change records reveal real capability. When a component becomes unavailable, engineers should evaluate electrical equivalence, software impact, layout, thermal limits, diagnostic behaviour and validation. A controlled successful substitution often demonstrates deeper ownership than a static origin statement.
Embedded software was part of the manufactured configuration
Software defines communication, diagnostics, control logic, calibration and fault response. Loading a binary at the end of assembly is an operation; designing, building, reviewing and sustaining the software is an engineering system. The two activities should earn different confidence.
The report said that from 1 June 2027 embedded software would need registration in an eligible Russian or Eurasian Economic Union register, or inclusion in a list for internal use. Producers should verify the final wording and route before treating a repository entry as sufficient.
Registration can establish administrative status, but it does not by itself demonstrate code quality, cybersecurity or vehicle compatibility. Evidence still needs requirements, architecture, static and dynamic analysis, unit and integration tests, vulnerability handling, signing, update control and rollback.
Hardware and software baselines must remain paired. A firmware version may support one sensor revision but not another; a bootloader update can change recovery; calibration can alter emissions, safety or performance. A score attached to a generic module family should not obscure configuration-specific evidence.
Surface mounting was a process capability
Local surface mounting places solder paste, chip components and other devices on a board before reflow. The visible operation is only part of capability. Stencil design, paste storage, placement accuracy, thermal profile, moisture control, inspection and rework determine yield and reliability.
Process qualification should define a stable window for representative board families. First-article inspection verifies setup, while ongoing statistical control detects drift. Automated optical inspection can broaden coverage, but its programmes require validation and false-call management.
Solder joints may pass visual inspection yet fail after temperature cycling or vibration. Appropriate destructive analysis, X-ray inspection, electrical test and environmental validation depend on package and risk. The control plan should connect detected defects to containment boundaries and root cause.
Rework deserves special governance. Excess heat or repeated handling can damage pads and components while leaving an apparently functional board. Limits, trained operators, traceability and post-rework tests protect the released configuration.
Additional points represented separable operations
Interfax said the draft would award additional points for software installation, enclosure assembly, functional testing and use of Russian components such as integrated circuits, printed circuit boards and connectors. Each activity adds value, but each answers a different dependency question.
Enclosure assembly can include sealing, thermal interface materials, connector retention and mechanical torque. Those operations directly affect ingress protection, heat transfer and vibration resistance. They require controlled materials and tools rather than a generic final assembly record.
Software installation needs binary identity, authorised keys, secure stations, successful verification and protection against loading the wrong variant. A flashing log should trace the installed image to the unit serial identity without exposing sensitive keys.
Functional testing checks specified inputs, outputs, communication and diagnostics. A short end-of-line cycle cannot reproduce every environmental and vehicle condition, so it should detect manufacturing escapes while qualification evidence establishes the broader design envelope.
Domestic components needed provenance and capacity
A component's origin claim should trace manufacturer, part number, production site where relevant, lot and conformity evidence. Distributors can preserve or weaken that chain. Counterfeit prevention, storage, moisture handling and date-code control remain necessary regardless of nationality.
Local component use increases resilience only if supply can meet volume, quality and lifecycle needs. A technically suitable chip with insufficient wafer, packaging or test capacity can become the new bottleneck. Programme planning should qualify output rate and recovery time, not merely sample availability.
Second sources improve resilience when they are truly qualified and independent. Two brands relying on one upstream die, substrate or packaging line may share a hidden failure domain. Supply mapping should extend to critical materials, tooling and test services.
Substitution rules need risk tiers. A passive component within an approved range may require limited verification; a processor, power device, sensor element or memory change can affect software timing, diagnostics, thermal behaviour and safety. Approval effort should follow consequence rather than purchase price.
Phased thresholds created an investment clock
For most electronic control units, the reported total threshold would be 205 points in 2026 and 2027 and rise to 245 from 2028. A staged requirement gives industry time to add capability, but it also creates a deadline portfolio across programmes with different launch dates.
A supplier should map every product against current points, mandatory gaps, planned investments, customer validation windows and the expected final rule. Equipment purchase alone may not close the gap if design rights, trained engineers, process qualification or component sources lag.
Projects need backward scheduling from the compliance date. Facility work, machine delivery, process development, sample builds, reliability testing, vehicle integration and documentary review consume sequential time. A late interpretation dispute can strand otherwise capable equipment.
Management should avoid optimising only for the threshold. A design narrowly arranged to collect points can become expensive to sustain. Investments should create reusable engineering and manufacturing platforms across module families where technically justified.
Sensors received a declining base-level bridge
The source described base and advanced requirement levels for electronic sensors. Producers meeting only the base level would receive 50% of the maximum possible norm before 2028, 25% before 2030, 15% before 2032 and no base-level points from 2032.
This declining credit is a transition signal, not proof that every sensor follows one technical path. Pressure, temperature, position, inertial, gas and optical sensors have different elements, packaging, calibration and test requirements. Category definitions in the adopted text will determine practical scope.
A supplier using the bridge should state the capability that replaces the lost credit at each date. Plans might include sensing-element control, packaging, application-specific circuitry, calibration automation, environmental testing or domestic inputs, depending on the final criteria.
The transition also affects customer programmes whose vehicle life extends beyond the regulatory step. A module qualified under an earlier level may face a later sourcing or recognition change. Contracts need responsibility for redesign, revalidation, inventory and service parts.
Points did not prove product safety
Localisation and safety overlap through control and traceability, but they are not interchangeable. A highly local module can still be unreliable; an imported component can still have strong evidence. Automotive release requires defined functional performance and risk-specific validation independent of the origin score.
Control units can influence braking, steering, powertrain, restraint, visibility or driver information. Safety analysis should identify hazards, safe states, diagnostic coverage and dependent failures. The rigour follows function, not the number of locally completed operations.
Cybersecurity belongs in the same lifecycle. External interfaces, diagnostic access, update mechanisms and supply-chain software create attack paths. Threat analysis, secure development, key management, vulnerability response and supported updates should accompany localisation.
Homologation and vehicle integration remain system responsibilities. A module can pass its bench specification yet interact badly with network timing, power sequencing, electromagnetic conditions or another controller. Vehicle-level tests close those interfaces.
Functional testing needed a layered strategy
Component incoming controls address identity and condition. Board tests examine assembly and electrical connectivity. Module end-of-line tests exercise interfaces and diagnostics. Environmental and life tests qualify the design, while vehicle tests validate interaction. Each layer catches a different failure class.
Test coverage should trace to requirements and failure analysis. Adding more test time without a coverage model may increase cost while leaving important faults invisible. Conversely, removing a test after yield improves requires evidence that upstream prevention and detection remain effective.
Golden units and test equipment require configuration control and calibration. A reference that drifts can approve bad products or reject good ones. Software in testers is part of the manufacturing system and needs change review like product software.
Failure data should flow back to design and process teams. Codes alone do not provide learning; engineers need defect location, component lot, machine, recipe, operator, test version and environmental history. Closed-loop analysis turns local testing into improving capability.
The vehicle maker and module supplier shared the boundary
The customer defines vehicle requirements, interfaces, validation and change approval; the supplier owns detailed design and process within the agreed boundary. Ambiguity creates either uncontrolled change or duplicated work. A responsibility matrix should include software, cybersecurity, components and field response.
Forecast stability affects economics. Electronics lines and component commitments require volume, while vehicle programmes can change mix. Contracts should address frozen periods, engineering changes, minimum quantities, obsolete material and liability for unapproved substitutions.
Field data completes qualification. Returns, diagnostic trouble codes, warranty repairs and software incidents need product genealogy. A score at initial recognition cannot substitute for continuing evidence from the fleet.
Economics depended on reusable capability
Local design and production add fixed cost before scale arrives. Business cases should separate one-time engineering, equipment, qualification, working capital, variable cost and expected service revenue. Public support or recognition can change incentives but cannot create customer volume by itself.
A platform approach can spread tools, process lines, software components and test infrastructure across related modules. Reuse must remain controlled: a common board or code library reduces cost only when variant requirements and regression tests are explicit.
Yield and first-pass test performance determine whether nominal local content becomes competitive serial output. Scrap, rework and long diagnostics consume imported and domestic inputs alike. Capability metrics should accompany point totals.
The dossier had to remain alive
A credible evidence dossier contains current rights, design baselines, bills of material, origin records, manufacturing routes, process qualifications, test results and change history. It should be generated from controlled systems where possible rather than assembled once before an audit.
Internal review should challenge traceability in both directions: from a claimed operation to production evidence and from a shipped serial unit back to its approved configuration. Sampling should cover ordinary shifts and suppliers, not only a prepared demonstration batch.
Exceptions need owners, expiry dates and containment. A temporary imported substitution, outsourced operation or emergency software build can be managed transparently; an undocumented exception destroys confidence in the score and the product.
Localisation became the ability to change responsibly
For automotive-electronics developers and vehicle manufacturers in Russia, the proposed ladder offered a more meaningful direction than counting final assembly alone. Rights, design, board manufacturing, software, testing and component provenance each reduced a different dependency.
The strongest programme will not treat the score as the last test. It will use mandatory foundations to secure product control, phased thresholds to schedule investment and additional operations to build serial competence. Safety, cybersecurity, quality and vehicle validation will remain separate release gates.
Points were therefore not parts and not proof of performance. They were a structured claim about where capability resided. The industrial achievement comes when that claim can survive a component shortage, a software defect, a field incident and a new vehicle generation without losing legal control, engineering understanding or production traceability.
HOT NEWS INTERNATIONAL
Leave a comment