The Year / Make / Model Problem
When a shop calls a yard and says "I need a 2019 Honda Accord door" — that description covers somewhere between four and eight distinct vehicle configurations depending on the trim level, engine, and production run. The door hinge points, mirror mount, power window module, and sensor cutouts can vary across those configurations.
Year, make, and model narrow down the pool. They don't identify the specific vehicle. For simple, undifferentiated parts — a generic rubber gasket, a cabin air filter — the distinction may not matter. For body panels with sensors, doors with integrated electronics, engines with multiple displacement variants, or transmissions tied to specific software calibrations, it matters considerably.
This is the core limitation of YMM-only sourcing: it assumes the part is generic across all vehicles sharing that year, make, and model. Some parts are. Many are not.
What a VIN Actually Encodes
The Vehicle Identification Number is a 17-character alphanumeric code standardized by NHTSA under 49 CFR Part 565, required on all vehicles manufactured for the US market since model year 1981. Each character position carries specific meaning.
The critical section for parts fitment is positions 4–8, the Vehicle Descriptor Section. Each manufacturer submits their own VDS encoding schema to NHTSA as part of their 565 regulatory submittal. There is no universal standard for what those five characters mean — Ford, Honda, Toyota, and BMW all encode different attributes in different positions within this section. This manufacturer-specific encoding is why VIN decoding requires a database (like NHTSA's vPIC), not just a formula.
For a 2019 Honda Accord, the VDS section encodes whether the vehicle has the 1.5L turbocharged four-cylinder or the 2.0L turbocharged four-cylinder, the body style, and the trim series. These differences are invisible from the year, make, and model alone.
Where Year / Make / Model Fails — Common Scenarios
Scenario 1: Engine Variants in the Same Model Year
2019 Honda Accord: Available with a 1.5L 4-cylinder turbocharged engine or a 2.0L 4-cylinder turbocharged engine. The oxygen sensors, engine mounts, motor mounts, exhaust manifold, and several related components are not interchangeable between these two variants.
A year/make/model search returns parts for "2019 Accord." A VIN search returns parts for the specific displacement in that vehicle.
Scenario 2: Mid-Year Production Changes
Manufacturers sometimes change suppliers, update component designs, or modify specifications during a production year without changing the model year designation. A 2017 pickup truck built in January and an identically-named truck built in September may use different brake calipers if the manufacturer switched suppliers mid-run. The VIN's production sequence (positions 12–17) can help identify which production batch a vehicle belongs to, allowing more precise fitment matching for affected components.
Scenario 3: Trim-Level Component Differences
A base-trim sedan and a sport-trim sedan of the same year and model often have different front fascias, different headlight assemblies (halogen vs LED), different suspension tuning, and different brake specifications. These parts are not interchangeable. The trim is encoded in the VDS section of the VIN — year/make/model alone doesn't surface it.
Scenario 4: Regional / Emissions Variants
The same model year vehicle may have been built to California Air Resources Board (CARB) emissions specs or federal emissions specs. Catalytic converters, oxygen sensors, and some engine management components differ between these configurations. The configuration is tied to the assembly plant and market — information encoded in the VIN but absent from a YMM query.
| What You Need to Know | YMM | VIN |
|---|---|---|
| Vehicle make, model, year | ✓ | ✓ |
| Engine displacement and configuration | Sometimes | ✓ (from VDS) |
| Trim level | Sometimes | ✓ (from VDS) |
| Drivetrain (FWD / AWD / 4WD) | Sometimes | ✓ (from VDS) |
| Assembly plant / market variant | ✗ | ✓ (position 11) |
| Production sequence / build date | ✗ | ✓ (positions 12–17) |
| Check digit validation (detects transcription errors) | ✗ | ✓ (position 9) |
How Radar Parts Uses VIN — The AI-First Approach
Radar Parts is built AI-first. That means VIN matching isn't a manual lookup step — it's a built-in layer of the sourcing logic, running in the background while the buyer sees a clean, minimal form. Here's how the system handles fitment, and how it will evolve:
Phase 1 — Current: Adaptive Confidence-Based Normalization
Buyer submits the request form
The form asks for year, make, model, and the part needed. VIN is optional — framed as an upgrade ("Have your VIN? It gets you more accurate results") rather than a requirement. The form is intentionally minimal: fewer fields means more submissions, and the normalization layer handles the resolution.
Claude Sonnet normalizes the request — in under two seconds
The submitted fields are passed to Claude Sonnet, which resolves the vehicle and part description against known configurations. If a VIN was provided, it's decoded via the NHTSA vPIC API first. Claude produces a normalized request — vehicle to make/model/trim/engine, part to standard nomenclature — and assigns a fitment confidence tier: HIGH, MEDIUM, or LOW.
GET https://vpic.nhtsa.dot.gov/api/vehicles/decodevin/{VIN}?format=json
// Returns 130+ attributes; key fitment fields:
"Trim": "Sport" // resolves trim ambiguity
"DisplacementL": "2.0" // resolves engine variant
"DriveType": "FWD"
"ErrorCode": "0" // 0 = clean decode
Confidence HIGH → sourcing begins, standard confirmation shown
When the normalization produces a clean, unambiguous vehicle-and-part spec — typically when a VIN is provided, or when the year/make/model/part combination has no meaningful configuration variants — the buyer sees a confirmation screen and sourcing begins immediately. "Your part is on our radar."
Confidence MEDIUM → one inline question, non-blocking
When normalization resolves the vehicle but finds one meaningful ambiguity — an engine variant that matters for the requested part, a trim line that affects fitment — the confirmation screen adds a single clarifying question before sourcing begins. The question is adaptive: Claude determines which single piece of information would most reduce ambiguity for this specific part type. For a transmission, that's engine displacement. For a door, that's trim level. For most mechanical parts, it's engine configuration. The buyer can answer or skip — but skipping is logged as unresolved ambiguity, and that flag carries through to offer results.
Confidence LOW → clarification screen replaces confirmation
When the normalization cannot resolve enough of the spec to source with reasonable confidence — missing trim on a highly variant model, ambiguous part name, or multiple engine options that produce non-interchangeable parts — the confirmation screen is replaced with a clarification screen. "We want to make sure we find the exact right part." The screen presents the specific question that matters most, with selectable options where possible (trim levels for that year/make/model, engine options, drivetrain) and a VIN entry field as the most precise resolution path. After the buyer responds, normalization re-runs with the new data and sourcing begins.
Example clarification questions (generated adaptively)
Trim ambiguity, body part: "Which trim level is your 2019 Camry?" → LE · SE · XSE · XLE · TRD · Hybrid · Not sure
Engine ambiguity, drivetrain part: "What engine does your Accord have?" → 1.5L 4-cylinder · 2.0L 4-cylinder · Not sure
Side ambiguity, body panel: "Is this for the driver side or passenger side?" → Driver (left) · Passenger (right)
VIN as fallback: "Have your VIN? It resolves all of these at once." → [VIN input field] · Skip and continue
Unresolved ambiguity flagged through to offer results
When a buyer selects "Not sure" or skips a clarification, the ambiguity isn't silently dropped — it's carried forward as a fitment confidence flag on the resulting offers. Results sourced against an unresolved trim ambiguity show a note: "Sourced against multiple trim configurations — verify fitment before accepting." This preserves the buyer's ability to move fast while ensuring they're not misled about the precision of the match.
Phase 2 — Roadmap: VIN-to-Interchange Mapping
The next evolution is mapping VIN-decoded vehicle attributes to industry interchange databases. The Auto Care Association's ACES (Aftermarket Catalog Exchange Standard) and the Vehicle Configuration Database (VCdb) define which parts fit which vehicles at the interchange level. Mapping a decoded VIN to ACES IDs allows the sourcing engine to search not just by part name, but by confirmed fitment interchange — surfacing parts from compatible vehicles, not just the exact same year/model, and eliminating parts that fit a related but different configuration.
This is how fitment confidence scoring evolves from "year/make/model match" to "VIN-confirmed interchange match" — a materially more accurate signal for the shop ordering the part.
Phase 3 — Roadmap: Supplier VIN-Tagged Inventory
Yards that maintain their inventory against the VIN of the donor vehicle — rather than just year/make/model — can provide a direct VIN-to-VIN match when the engine or body configuration aligns. As the Radar Parts supplier network grows, yards with VIN-tagged inventory will be scored higher for fitment confidence on complex part types, creating a marketplace incentive for better supplier data.
Data Sources Behind VIN Fitment
NHTSA vPIC API — The primary free source for VIN decoding. Provided by the US Department of Transportation, free to use without registration, updated with manufacturer 565 submittals. Coverage is approximately 99% accurate for model year 1995 and forward for major manufacturers, with declining confidence for 1980–1994 vehicles. Prior to 1980, VINs were not standardized and vPIC does not decode them. Available as a live API (rate limited) or as a downloadable standalone PostgreSQL / SQL Server database for high-volume applications.
ACES / VCdb — The Auto Care Association's Aftermarket Catalog Exchange Standard and Vehicle Configuration Database. This is the professional standard for parts-to-vehicle fitment mapping used by aftermarket parts manufacturers, distributors, and retailers. Licensing required for commercial use; covers fitment at the sub-model and engine-configuration level. This is the data layer that enables the "does part X fit vehicle Y" lookup in commercial parts platforms.
DataOne / other VIN explode services — Commercial providers that extend beyond NHTSA's vPIC to provide "VIN explosion" — additional vehicle attributes beyond what the VIN pattern encodes, including optional equipment, packages, and feature-level configuration from manufacturer build data. Useful for trim-sensitive parts where the VDS section alone doesn't resolve the configuration fully.
Car-Part.com interchange data — The salvage industry's primary interchange reference, which maps parts fitment across years, makes, and models for recycled parts. When a buyer provides a VIN, the resolved year/make/model/engine can be used to query Car-Part.com's interchange layer to find compatible parts from related vehicles — expanding sourcing options beyond exact-match results.
Limitations of VIN-Based Fitment
VIN decoding is accurate but not perfect. Several limitations apply:
- VDS is manufacturer-specific and not always complete. Some manufacturers encode only a subset of vehicle attributes in the VDS. A VIN may decode to year/make/model/engine but not resolve trim fully — particularly for older vehicles or niche models.
- Running production changes aren't always reflected. A mid-year specification change at the assembly plant level isn't always captured in the VIN pattern. For affected part types, production date range (derived from the VIS sequence) becomes the relevant filter.
- Pre-1981 vehicles lack standardized VINs. Vehicles built before NHTSA's 1981 mandate use manufacturer-specific formats that vPIC cannot decode. Parts sourcing for these vehicles typically requires direct communication with the yard.
- VIN alone doesn't guarantee physical fitment. A VIN confirms the vehicle's documented configuration. It doesn't account for previous modifications, non-factory equipment, or post-production changes to the vehicle. Visual inspection and OEM part number cross-reference remain the final verification step for precision-fit applications.
Best Practice: How to Source with Confidence
For the highest fitment confidence on any part order:
- Provide the full 17-digit VIN whenever possible. It's the single most accurate vehicle identifier available and enables engine, trim, and plant-level matching that YMM alone cannot.
- Note the OEM part number from the vehicle's existing component if accessible — either stamped on the part, in the owner's documentation, or via a dealer parts lookup using the VIN. OEM part number cross-reference is the highest-confidence fitment verification method.
- Specify trim level when the VIN isn't available. Base, Sport, EX, Limited, and similar trim designations resolve many of the configuration ambiguities that bare year/make/model leaves open.
- Flag electronically-integrated parts explicitly. ECUs, TCMs, ABS modules, airbag controllers, and ADAS sensors may require programming to the vehicle's VIN after installation. This is a separate consideration from physical fitment — and worth surfacing in the sourcing request so the yard can flag whether the part is programmable or paired to a specific VIN.
Submit with Your VIN — We'll Handle the Rest
Radar Parts accepts VIN as part of the request form. Our sourcing engine decodes it, resolves the spec, and uses it to match against our yard network with higher fitment confidence than year/make/model alone.
Request Early Access →