A Digital Circularity Vehicle Passport built on Midnight's zero-knowledge blockchain.
Prove what a vehicle is. Reveal only what the asker is entitled to see.
Status: pre-alpha. Nothing here is production software. This repository is being rebuilt from scratch as of August 2026 on top of ODATANO/NIGHTGATE. Read
docs/DECISIONS.mdbefore contributing — several architectural choices are settled and deliberately not open for re-litigation.
From 1 September 2032, every vehicle placed on the EU market must carry a Digital Circularity Vehicle Passport under Regulation (EU) 2026/1738, which entered into force on 13 August 2026. The regulation requires that passport to be "aligned, interoperable and, where possible, integrated with other vehicle related environmental passports established under Union law" — most obviously the EV battery passport that becomes mandatory in February 2027 under Regulation (EU) 2023/1542.
ShieldVIN implements that vehicle passport with selective disclosure: one canonical vehicle record, anchored once, from which each party — buyer, dealer, recycler, insurer, regulator — can be shown exactly what they are entitled to and provably nothing more.
The compliance obligation is the reason the passport exists. The reason it is worth building privately is everything downstream of it: odometer fraud, title washing, undisclosed accident history, and the ordinary asymmetry where a used-car buyer has no way to verify what they are told.
A vehicle record is canonicalised, hashed, and anchored on Midnight. A parallel Merkle tree commits to individual provable fields. From then on:
- Anchoring proves the record existed in a specific state at a specific time.
- Predicate proofs answer questions like "is the odometer above 150,000 km?" or "has this vehicle ever been written off?" — returning a yes or no that is verified on-chain, without ever revealing the underlying value.
- Disclosure grants control who may read which tier, enforced both in the API and on-chain.
The hard part is not the cryptography — NIGHTGATE already
provides it. The hard part is the honest one, described plainly in
docs/ARCHITECTURE.md: proving a number is the anchored number is not the
same as proving it is true. Closing that gap is what distinguishes this project, and it is
deliberately scoped as Phase 2 rather than pretended away in Phase 0.
| NIGHTGATE | OData V4 gateway to Midnight — anchoring, ZK predicates, disclosure grants |
| Midnight Network | Zero-knowledge blockchain; ledger 8, Compact 0.31.x |
| SAP CAP | Application framework — CDS domain model, OData services |
ShieldVIN is a consumer application of NIGHTGATE, in the same relation to it as NIGHTPASS is for battery passports. We do not fork it, and we contribute upstream where our needs are general rather than vehicle-specific.
| Document | What it covers |
|---|---|
| ARCHITECTURE.md | System design, the trust model, and its honest limits |
| DECISIONS.md | Settled decisions and the reasoning behind them |
| ROADMAP.md | Phased delivery plan |
| FIELDS.md | The provable-field registry and why it is capacity-constrained |
Apache-2.0, matching the ODATANO stack.