# Optimistic Verifiable Claims: A Blockchain Protocol for Conditionally Confidential Bidding in Decentralized Manufacturing

Crypto.world

> In decentralized manufacturing, a Provider cannot price a job accurately without inspecting the design file (such as G-code), but the Consumer cannot share that file before a contract exists without…

**arXiv**: 2607.25517

**Published**: 2026-07-28

**Authors**: Marko Corn, Nejc Rožman, Primož Podržaj

## AI summary

## Overview

**Research area:** Blockchain protocol design for decentralized manufacturing — specifically applied cryptography, smart-contract engineering, and pre-contractual market design for additive manufacturing.

**Technical level:** Advanced. The paper formalizes a cryptographic primitive (commitment, private witness, deterministic verification function) as a finite-state machine, implements four verifiers in Solidity in both monolithic and segmented execution strategies, and reports gas-derived cost and settlement-time measurements across three production chains.

**Scope (one sentence):** The paper introduces and empirically evaluates the Optimistic Verifiable Claim (OVC), a blockchain protocol that lets a Consumer commit to claims about a concealed design file so a Provider can price and bid on it without seeing it, with honesty enforced only if the claim is challenged.

## What This Paper Is About

In decentralized manufacturing, a Provider cannot price a job accurately without inspecting the design file (such as G-code), but the Consumer cannot share that file before a contract exists without exposing intellectual property — a "deadlock of trust" the authors describe as a pre-contractual instance of Arrow's information paradox. Existing tools solve only half the problem: access-control schemes limit who can read a file but say nothing about whether the file matches the claims used for pricing, and escrow or fair-exchange protocols guarantee delivery but are content-agnostic.

The goal is a protocol that makes a claim about a hidden artifact binding and provably falsifiable without a trusted intermediary: the claim is accepted by default, the design stays private on the honest path, and a challenge triggers a deterministic on-chain check that exposes dishonesty.

## Key Contributions

1. **A reframing of the pre-contractual trust deadlock.** Rather than trying to verify a claim before contracting (which the deadlock makes impossible), the authors recast confidential bidding as the optimistic adjudication of committed claims, where certain exposure under dispute — not pre-bid verification or a trusted broker — is what makes a claim usable for pricing.
2. **The OVC primitive and a four-stage manufacturing pipeline.** OVC is formalized as a commitment, a private witness, and a deterministic on-chain verification function with precise fault attribution, then instantiated four times (Access, Identity, Conformance, Feature) into a pipeline whose cross-instance consistency is enforced at bind time. All four are implemented in Solidity in both monolithic and segmented execution strategies.
3. **An empirical characterization of the "single-transaction ceiling."** The authors identify the artifact size beyond which one-shot verification exceeds the per-transaction gas limit of every real chain, and show that partitioning the work across bounded transactions restores feasibility up to 50 MB artifacts while relocating, not reducing, the O(n) work.
4. **A deployment analysis across the scalability trilemma.** Measured gas is projected onto Ethereum, Arbitrum, and opBNB, quantifying settlement cost and dispute duration at industrial artifact sizes and showing that execution-layer choice, not protocol design, bounds practicality.

## Main Findings

- **3DBenchy baseline (6.41 MB, 173,351 lines of G-code, PLA at 0.15 mm layer height, Prusa MK3):** The no-dispute outcome costs $7,207 in up to 9 hours on Ethereum, $288 in 3 min on Arbitrum, and $2.87 in 2 min on opBNB. A fully contested dispute costs $49,660 in up to 57 hours on Ethereum, $1,988 in 19 min on Arbitrum, and $19.73 in 13 min on opBNB.
- **Industrial-scale projection (50 MB design):** An undisputed service reaches $56,173 and up to 3 days on Ethereum against $22.36 and 16 min on opBNB. A fully contested dispute reaches $488,440 over up to 18 days on Ethereum against $195 and 1.6 hours on opBNB.
- **The semantic check dominates cost:** Of the four verifiers, the material-consumption check (OVC-Feature) is the costliest, its predicate being the most expensive to evaluate on-chain. This is why the semantic predicate, rather than the cryptographic hash check, becomes the limiting constraint on segmentation.
- **Chain choice, not protocol design, is the binding constraint:** OVC makes confidential, claim-based bidding economically feasible on Arbitrum and opBNB, but not on Ethereum at industrial scale.
- **Cost structure is asymmetric by phase:** The O(N) artifact upload is paid once at bidding, independent of how many Providers bid, while binding to a chosen Provider is O(1) — a 32-byte key encrypted under the Provider's public key.
- **Segmentation has a known price:** Partitioning verification across bounded transactions restores feasibility up to 50 MB artifacts but relocates the necessary O(n) work rather than reducing it.
- **Confidentiality is optimistic, not cryptographic hiding.** The construction c = H(w) provides computational binding under collision resistance but does not provide information-theoretic hiding; a computationally unbounded party who knows c can in principle brute-force w if the witness space is small. Pre-dispute confidentiality is a protocol property — the witness is never published unless a dispute is triggered and the Consumer elects to disclose.
- **Disputes are the deliberate confidentiality exception.** Proving a contested claim requires the Consumer to disclose the witness on-chain. A counterparty set on forcing disclosure can do so, at which point the Consumer's recourse is to withhold, resolving to Withheld and forgoing the contract to keep the artifact secret.
- **Fault attribution is four-way:** VALID if verification passes, INCONSISTENT if H(w) ≠ c, INCORRECT if the hash matches but the semantic predicate fails, and WITHHELD if w is not disclosed within the configurable dispute window W_d.
- **Cross-instance substitution is a real attack the design closes.** If each instance were committed in independent transactions, a dishonest Consumer could register a different artifact hash to OVC-Conformance than to OVC-Identity, satisfying the conformance check with an unrelated file. The BindingPhase coordinator transaction reads the hashes from the sealed Demand and propagates them identically to all four instances, making consistency a contract invariant rather than an off-chain assumption.
- **Not reported in the provided text:** the per-transaction gas limit thresholds, chunk sizes, absolute gas figures, and the concrete definition of the feature-extraction function (referenced to Section 3.6.1) are not included in the excerpt.

## Methodology in Plain English

The authors start from a simple observation: you cannot check a claim about a secret without seeing the secret. So instead of trying to verify before contracting, they bind the claim up front and make it expensive to lie.

Each claim is a triple: a public commitment, a private witness held by the Consumer, and a deterministic verification function run by the smart contract alone — no off-chain computation and no trusted oracles. The commitment is a collision-resistant hash of the witness. Verification has two parts joined by AND: does the revealed witness hash to the commitment, and does the witness satisfy the specific semantic property being claimed (valid key linkage, consistent delivery, syntactically valid G-code, or the declared material consumption).

The protocol is a finite-state machine. A claim moves from COMMITTED to ACCEPTED if the Provider simply accepts, with no data disclosure. Only a challenge diverts it to adjudication, where the Consumer discloses the witness on-chain and the contract returns a verdict. A dispute window measured in blocks guarantees liveness: if the witness is never disclosed, anyone can resolve the claim as Withheld. The window is a deployment parameter — a few thousand blocks on fast-settling L2s, tens of thousands on L1 — and does not affect verification logic.

Four instances are instantiated in one atomic transaction at binding: OVC-Access (commitment to the hash of the XOR decryption key plus a Provider-linkage commitment), OVC-Identity (hash of key, plaintext, and ciphertext, plus a check that G ⊕ K = C), OVC-Conformance (artifact hash plus a syntactic G-code validator checking for standard motion commands such as G0 and G1), and OVC-Feature (artifact hash plus an on-chain feature extraction f(G), here net material consumption). The Demand is posted at bidding and carries the artifact hash, the claimed feature value, and the encrypted artifact uploaded on-chain, so delivery itself becomes adjudicable. Because the ciphertext is already on the ledger, binding writes only O(1) state regardless of file size.

For measurement, the authors run these checks on a real 6.41 MB G-code file (the 3DBenchy) and implement each in Solidity in two strategies — monolithic (single transaction) and segmented (bounded chunked transactions) — then project the measured gas onto Ethereum, Arbitrum, and opBNB to produce dollar costs and settlement times at both 6.41 MB and a projected 50 MB industrial artifact.

## Why This Matters

**Impact on research.** The paper offers a third path between two dominant approaches: access control, which protects confidentiality but proves nothing about content, and zero-knowledge proofs, which provide strong semantics but carry proving overhead and integration complexity that the authors argue is hard to absorb in latency-sensitive industrial bidding. By shifting from pre-contract verification to post-commitment adjudication, it turns an impossible verification problem into an economic one — lying is safe only until challenged. The paper also contributes a concrete empirical result: the point at which on-chain adjudication stops fitting in a single transaction, and what segmenting it costs.

**Real-world applications:**

- **Additive manufacturing bureaus and print-on-demand networks** pricing jobs from declared material consumption without receiving the customer's design file.
- **IP-sensitive design marketplaces** where proprietary geometries and process-specific toolpaths must be quoted before a contract exists.
- **Capacity reservation and quoting in distributed production networks**, where the O(N) upload is amortized across every Provider that bids and binding a winner stays O(1).
- **Auditable dispute records for industrial marketplaces**, where the on-chain verdict with fault attribution can be consumed as evidence for penalties or reputation scoring.

**Industry relevance.** The headline result is operational, not theoretical: the same protocol is practical on Arbitrum and opBNB and impractical on Ethereum at industrial scale, so deployment architecture — execution-layer choice — determines viability. For shop-floor adoption, the OVC adjudicates outcomes rather than settling them, deliberately leaving economic and social consequences to the surrounding marketplace, which means it can be layered onto existing platforms rather than replacing them.

## Future Directions

- **Zero-knowledge adjudication for disclosure-free confidentiality.** The authors name this as the direct extension: proving a contested claim without publishing the witness on-chain, which would remove the current trade-off where a counterparty can force disclosure and the Consumer's only recourse is to withhold and forgo the contract.
- **Stronger predicates beyond lightweight checks.** The four verifiers are intentionally O(n) and each checks one specific property rather than general manufacturability. Extending toward richer semantic guarantees raises the open question of how much semantic work can fit inside gas limits.
- **Parameters that adapt to chain and artifact size.** The dispute window W_d is a deployment parameter whose appropriate value depends on block time, artifact size, and SLA requirements; chunk size and window selection at 50 MB scale are left as deployment questions.
- **Marketplace-level integration of verdicts.** Since OVC adjudicates rather than settles, the design leaves open how reputation systems, penalties, and staking should consume the four-way outcome record.

## Target Audience

Researchers and engineers working at the intersection of blockchain and manufacturing systems — particularly those designing decentralized marketplaces, smart-contract protocols for industrial coordination, or IP-protection mechanisms for additive manufacturing. It will also be useful to practitioners evaluating which chain to deploy an on-chain adjudication scheme on, and to cryptography researchers interested in the trade-off between optimistic disclosure and zero-knowledge approaches to confidential bidding. Readers need comfort with commitment schemes, hash-based binding, and smart-contract gas economics; the paper assumes a research-level audience rather than a beginner one.

## Authors’ abstract

Decentralized manufacturing faces a pre-contractual impasse: a Provider cannot price a service accurately without inspecting the design file, yet the Consumer cannot share that file without exposing intellectual property. We introduce the Optimistic Verifiable Claim (OVC), a blockchain protocol that lets a Consumer publish a verifiable claim about a concealed design (such as the material it consumes) and a Provider price and bid on it without seeing the design. The claim is…

[The full abstract on arXiv](https://arxiv.org/abs/2607.25517)

[Read the original paper](https://arxiv.org/abs/2607.25517)

---

Canonical: https://crypto.world/paper/optimistic-verifiable-claims-a-blockchain-protocol-for-conditionally-2607.25517
