EV-29 — Real Vehicle Examples

ASSUMPTION — This lesson teaches how to read and evaluate real vehicle data responsibly; it does not fabricate specific model numbers. Actual vehicle data is kept verified with current sources in data/vehicles/, separate from this academy content.

1. Why Verification Matters

Not every technical fact about a production vehicle is publicly known. OEMs routinely disclose headline figures — battery capacity, peak power, official range — through press materials and homologation documents, but they do not always disclose cell chemistry, cell supplier, exact pack architecture, or software calibration details. Treating an unverified number as if it were confirmed is worse than saying nothing, because it misleads the reader into false confidence, and once a wrong number propagates it is hard to correct. This is precisely the trap this academy avoids: rather than filling a gap with a plausible-sounding figure, unverifiable information is either labeled by its actual confidence level or explicitly marked “Publicly undisclosed.”

FACT — A confident-sounding number is not the same as a correct number. The discipline of labeling confidence exists specifically because readers cannot tell the difference between a verified fact and a well-phrased guess unless the source tells them.

2. Verification Labels

Every piece of vehicle-specific data in a well-run reference system should carry one of these labels:

Label	Meaning	Typical source
OFFICIALLY_DISCLOSED	Stated directly by the OEM or a regulatory/homologation authority	Press kit, spec sheet, type-approval filing
REPORTED	Stated by a reputable independent press outlet, not directly by the OEM	Trade journal, established automotive press
TEARDOWN	Determined by physically disassembling and analyzing a real unit	Independent engineering teardown report
INFERRED	Deduced indirectly from other known facts, not measured or stated directly	Calculation from disclosed specs, patent filings
ESTIMATED	A best-effort approximation with no direct source	Analyst estimate, extrapolation

FACT — LOW-confidence labels (INFERRED, ESTIMATED, and to a lesser extent TEARDOWN) are not hidden from the reader or silently upgraded to sound more certain — the label travels with the number everywhere it is displayed, so the reader always knows how much weight to give it.

3. The Vehicle Data Model

A vehicle record in a proper reference system captures more than a single spec sheet — it tracks where each field came from and when it was last checked. The typical fields are: brand, model, model_year, market, vehicle_type, platform, battery (capacity, chemistry, format), motor (type, drive, config), charging (ac, dc, voltage), thermal, range, efficiency, verification_status, confidence, source, last_verified.

The last three fields — verification_status, confidence, and source — are what separate a trustworthy reference from a marketing spec sheet reprint: they let a reader trace any number back to where it came from and how reliable that origin actually is, and last_verified flags data that may be stale because specifications, trims, and even chemistry can change across a model’s production run without a new model year being announced.

4. Worked Examples of Reading Confidence

EXAMPLE — Claim: “This model uses an 800V architecture.” If the OEM’s own technical materials state this directly, the correct label is OFFICIALLY_DISCLOSED, and a reader can treat it as fact. If instead an independent teardown measured the pack’s series configuration and calculated the nominal voltage from it, the correct label is TEARDOWN or INFERRED — still useful, but it should be read as “very likely true based on physical evidence,” not “confirmed by the manufacturer.”

EXAMPLE — Claim: “Pack voltage may be roughly 700–800V.” Phrasing it as a range with “may be” is itself doing labeling work — this is an ESTIMATE, and the range width should roughly reflect how uncertain the underlying reasoning actually is. A narrow range presented with hedging language but no real basis is more misleading than a wide range that honestly reflects the uncertainty.

INTERPRETATION — When a claim’s source is not stated at all, the safest reader behavior is to treat it as unverified regardless of how confidently it is phrased — confident phrasing is not evidence.

5. The Limits of Teardown Data

Teardown analysis is often treated as the gold standard of independent verification, and it is genuinely valuable — it produces physical, measurable facts rather than marketing claims. But it has real limits that a careful reader should keep in mind. A teardown examines one physical unit (or a small number of units), so anything that varies across a production run — supplier changes mid-cycle, regional trim differences, running production changes — may not be captured by a single sample. A teardown also reflects the vehicle at one point in time; if the OEM updates a component later in the production run, older teardown data becomes outdated without necessarily being wrong about the unit it examined.

FACT — This is exactly why TEARDOWN is its own label rather than being folded into OFFICIALLY_DISCLOSED: it is strong physical evidence, but “true for the unit examined” is not automatically the same claim as “true for every unit of this model ever produced.”

6. FAQ

Why is it correct to say “Publicly undisclosed”?

FACT — OEMs do not always disclose battery chemistry, cell supplier, or other technical details; when no reliable source states a fact, saying so plainly is more accurate and more useful to the reader than presenting an unverified guess as if it were confirmed.

Is teardown data reliable?

FACT — Teardown/analysis data is concrete physical evidence, but it is sample-limited and time-limited, so it should carry its own label (generally treated as lower confidence than an official OEM statement) rather than being presented as equivalent to it.

If a number appears on many websites, does that make it verified?

INTERPRETATION — Not necessarily — an unsourced number can be copied from site to site until it looks authoritative purely through repetition. Verification should trace back to an original source (OEM statement, homologation document, or a specific teardown), not to how many places repeat the same figure.

7. Summary

• If a fact cannot be traced to a reliable source, the honest answer is “Publicly undisclosed” — not a plausible-sounding guess.
• The five labels (OFFICIALLY_DISCLOSED, REPORTED, TEARDOWN, INFERRED, ESTIMATED) form a confidence ladder, and the label should travel with the number wherever it is shown.
• The vehicle data model’s verification_status, confidence, and source fields are what make a reference trustworthy rather than just a repost of marketing copy.
• Teardown evidence is physical and concrete but limited to the unit(s) examined and the point in time examined — it is strong evidence, not automatic proof for an entire production run.
• Confident phrasing is not evidence; only a traceable source is.

8. Sources and Verification Note

No model-specific hard technical data is given in this lesson — only the methodology for evaluating and labeling such data. Actual vehicle records are maintained separately with current sources.

• OEM documentation (general) — press kits, spec sheets, homologation filings as source-type examples.
• SAE J1715 — Hybrid and electric vehicle terminology.
• SAE J2954 — referenced conceptually as an example of how a technical standard document itself functions as an OFFICIALLY_DISCLOSED-tier source for platform-level claims.

ASSUMPTION — Any real vehicle data referenced elsewhere on the site must carry its own label and be re-verified before publication; nothing in this lesson should be read as a claim about a specific model.

Next Lesson

• EV-30 — EV Buying Guide: decision criteria for choosing an electric vehicle.