CARVE-Q: A Quantum-Classical Architecture for Certified Autonomous Driving Maneuver Repair
By
[Submitted on 3 Jun 2026]
Not artisan, but a perfectly fine bagel. Hits the spot.
Summary
This paper introduces CARVE-Q, a hybrid quantum-classical architecture for certifying safe driving maneuver repairs after a veto. CARVE (Certified Affordable Repair of Vetoed maneuvers via Envelopes) constructs a finite repair lattice and emits structured certificates recording binding rules, cooperation envelopes, and responsibility-weighted cost splits. The key bottleneck is that multi-owner repair induces a product lattice requiring classical minimum finding in Θ(M) queries. CARVE-Q addresses this by applying quantum minimum finding (Durr-Hoyer/Grover) to the black-box lattice, achieving O(√M) oracle queries while keeping all safety authority classical. The system is validated on Lanelet2-grounded INTERACTION replay with 100% right-of-way respect, 100% blame consistency, and zero priority false positives, demonstrating a trust-bounded quantum-AI pattern where quantum proposes and CARVE certifies.
Key quotes
· 5 pulledThe critical question after a correct driving veto is not only whether a maneuver is unsafe, but whether the blocked interaction admits a lawful, auditable, and responsibility-bounded repair.
We introduce CARVE, Certified Affordable Repair of Vetoed maneuvers via Envelopes, a certificate architecture for prediction-free interactive repair.
In the conservative verifier-oracle model, exact classical minimum finding requires Θ(M) queries in the worst case, whereas Durr-Hoyer/Grover minimum finding uses O(√M) oracle queries with high probability.
The result is a trust-bounded quantum-AI pattern for certified autonomy: quantum proposes; CARVE certifies.
We demonstrate state-vector minimum finding on CARVE repair oracles up to 65,536 assignments and validate certificate preservation on Lanelet2-grounded INTERACTION replay with 100% right-of-way respect, 100% blame consistency, and zero priority false positives.
You might also wanna read
New Benchmark Reveals High Rates of Outcome-Driven Constraint Violations in Autonomous AI Agents
Researchers introduce a new benchmark for evaluating autonomous AI agents' safety, specifically focusing on outcome-driven constraint violat
arxiv.org·3mo ago
Testing Quantum Computing Claims: Replacing IBM QPU with /dev/urandom in ECDLP Attack
This article tests a quantum computing claim from the Q-Day Prize submission by replacing the IBM Quantum Processing Unit (QPU) with /dev/ur
github.com·1mo ago
Claude Code Debugs Low-Level Bug in Post-Quantum Cryptography Implementation
The author describes their experience implementing ML-DSA, a post-quantum signature algorithm, in Go over four days. Despite completing the
Sublinear-Space ZKP System in Rust: Streaming Prover with O(√T) Memory via KZG Commitments
A reference implementation of a sublinear-space zero-knowledge proof (ZKP) prover/verifier system in Rust, based on a whitepaper. The stream
github.com·8mo ago
Formal Framework for LLM-Verifier Systems: Convergence Theorem and 4/δ Latency Bound
This research paper presents a formal framework for integrating Large Language Models with Formal Verification tools, addressing reliability
arxiv.org·5mo ago
Demonstration of Algorithmic Quantum Speedup for an Abelian Hidden Subgroup Problem
journals.aps.org·11mo ago