{"schema_version":1,"research_id":"eoa_inverse_innovation_exp03_external48_20260801","source_assessment_id":"eoa_inverse_innovation_exp03_opportunity320_20260801","cell_id":"computability_boundary_mapping__architecture_urban_planning","selection_stratum":"DEPLOYABLE_PRIORITY","search_queries":["site:gov.uk automated building regulations compliance checking BIM report","automated compliance checking building codes BIM primary research decidability rule language","buildingSMART IDS specification automatic compliance checking official","Singapore CORENET X automated model checker building regulatory approval official","site:doi.org automated building code compliance checking limitations ambiguity computable rules BIM","site:nist.gov building code automated compliance checking BIM rules","automated building code compliance checking rule language expressiveness formal verification termination","site:info.corenet.gov.sg \"Automated Model Checker\" limitations results QP responsibility","site:support.corenet.gov.sg model checker check results regulatory approval not replace review","Turing 1936 computable numbers Entscheidungsproblem original paper DOI halting circle-free machine","site:lean-lang.org termination proof documentation structural recursion","automated building compliance checker \"UNKNOWN\" pass fail not applicable","digital building permit automated checker decision support human in loop prior art","site:bls.gov/ooh software developers architects median pay 2024"],"sources":[{"source_id":"S1","title":"On Computable Numbers, with an Application to the Entscheidungsproblem","publisher":"Proceedings of the London Mathematical Society / Wiley","url":"https://doi.org/10.1112/plms/s2-42.1.230","source_class":"PRIMARY_RESEARCH","publication_date":"1937","accessed_at":"2026-08-02","claims_supported":["Turing establishes that no general finite decision procedure exists for the relevant unrestricted machine problem.","This general computability result applies to the candidate only if the accepted plan-rule contract admits an answer-preserving encoding of unrestricted computation; the paper alone does not establish that domain premise."]},{"source_id":"S2","title":"Recursive Definitions","publisher":"Lean Language Reference","url":"https://lean-lang.org/doc/reference/latest/Definitions/Recursive-Definitions/","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"2026","accessed_at":"2026-08-02","claims_supported":["Structural and well-founded recursion provide enforceable ways to admit recursive definitions with termination evidence.","Automatic termination analysis is intentionally incomplete, and some terminating definitions require explicit measures or proofs.","A proof-carrying total rule fragment is technically plausible."]},{"source_id":"S3","title":"Digital approaches to construction compliance checking: Validating the suitability of an ecosystem approach to compliance checking","publisher":"Advanced Engineering Informatics / Elsevier","url":"https://orca.cardiff.ac.uk/id/eprint/164367/1/1-s2.0-S1474034623004160-main.pdf","source_class":"PRIMARY_RESEARCH","publication_date":"2023-11-24","accessed_at":"2026-08-02","claims_supported":["A prototype extensible compliance ecosystem already returned true, false, or unknown, tried alternate checking services, and routed unresolved cases to human decision.","The prototype retained manual oversight and audit records and compared automated results with a qualified professional's manual assessment.","Digitizing three regulatory documents required 83, 164, and 143 labor-hours respectively.","The one-building demonstration reported 102 minutes for automated checking versus 22 hours for manual checking, but does not establish general performance.","Consultation participants identified BIM adoption, data requirements, consistency, accuracy, and role degradation as implementation concerns."]},{"source_id":"S4","title":"Information Delivery Specification developer guide","publisher":"buildingSMART International","url":"https://github.com/buildingSMART/IDS/blob/development/Documentation/ImplementersDocumentation/developer-guide.md","source_class":"STANDARD","publication_date":"2024","accessed_at":"2026-08-02","claims_supported":["IDS defines a constrained XML/XSD-based information-requirement language with conformance test cases.","Implementations are expected to disclose an inability to parse a nominated IFC version.","The guide restricts allowable constructs, distinguishes required, optional, and prohibited specifications, and requires checker conformance to a test suite.","This is concrete prior art for enforceable syntax boundaries and limitation disclosure, although IDS is not a universal building-code decision language."]},{"source_id":"S5","title":"What is MC?","publisher":"CORENET X, Government of Singapore","url":"https://support.corenet.gov.sg/hc/en-us/articles/13750775323791-What-is-MC","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"2026-06-02","accessed_at":"2026-08-02","claims_supported":["CORENET X's Model Checker is officially described as a supplementary tool.","The current beta automatically checks IFC+SG models against basic IFC standards, modelling standards, and selected regulatory requirements rather than promising universal compliance decisions."]},{"source_id":"S6","title":"CORENET X: Mandatory submissions starting from 1 October 2025","publisher":"Urban Redevelopment Authority, Government of Singapore","url":"https://www.ura.gov.sg/guidelines/circulars/dc25-07/","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"2025-09-10","accessed_at":"2026-08-02","claims_supported":["CORENET X is an official integrated building-approval process involving multiple regulatory agencies.","More than 50 voluntary pilot projects had been onboarded before phased mandatory submission began.","The program evidences regulator and industry demand for digital permitting infrastructure, training, openBIM submissions, and model-quality support, but not specifically for computability-boundary proofs."]},{"source_id":"S7","title":"Building Control Independent Panel report","publisher":"UK Government","url":"https://www.gov.uk/government/publications/building-control-independent-panel-report/building-control-independent-panel-report","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"2026-05-20","accessed_at":"2026-08-02","claims_supported":["Building control is a public-safety function involving compliance decisions based on plans, evidence, inspections, and professional judgment.","The report calls for shared digital infrastructure, common data standards, transparent reporting, and visible accountability.","It identifies fragmentation, uneven capacity, inconsistent oversight, and resource pressure while preserving the responsibilities of regulators and dutyholders."]},{"source_id":"S8","title":"Architects","publisher":"U.S. Bureau of Labor Statistics","url":"https://www.bls.gov/ooh/architecture-and-engineering/architects.htm","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"2025-08-28","accessed_at":"2026-08-02","claims_supported":["The May 2024 median annual wage for architects was $96,690.","Licensed architectural and code-domain labor is a material cost component of formalizing and validating building rules."]}],"problem_evidence":{"support":"WEAK","rationale":"The surrounding problem is externally recognizable: compliance review is safety-significant, rule digitization is labor-intensive, and real checkers deliberately limit themselves to selected requirements or preserve unknown and human-review states (S3, S5, S7). However, the bounded search found no authority or deployed product actually promising a total, exact Boolean decision for every plan in an extensible rule language, and no evidence that timeouts are routinely collapsed into legal pass/fail outcomes. The candidate's exact prevalence premise therefore remains unverified.","source_ids":["S3","S5","S7"]},"stakeholder_evidence":{"support":"MODERATE","rationale":"Regulatory pull for digital permitting, model checking, common data standards, and accountable workflows is concrete: Singapore has phased mandatory CORENET X submissions after more than 50 pilots, while the UK panel recommends a digital building-control backbone (S6, S7). This does not establish procurement demand for the narrower computability-proof intervention.","source_ids":["S6","S7"]},"prior_art":{"proximity":"SUBSTANTIAL_COLLISION","closest_analogues":[{"name":"D-COM ecosystem of compliance-checking services","similarity":"This implemented research prototype is extensible, routes checks among services, returns true/false/unknown, escalates unresolved results to human decision, retains manual oversight, and records an audit trail—covering much of the candidate's labeled-fallback and authority-preserving architecture.","remaining_difference":"It does not report a formal classification of rule-language expressiveness, a proof that admitted fragments are total and exact, an impossibility reduction for unrestricted rules, or routing controlled by machine-enforced proof certificates and declared computational bounds.","source_ids":["S3"]},{"name":"buildingSMART Information Delivery Specification","similarity":"IDS supplies a constrained machine-readable language, schema validation, conformance tests, restricted constructs, and disclosure of unsupported IFC versions, demonstrating enforceable scope boundaries and testable checker contracts.","remaining_difference":"IDS primarily checks alphanumeric information-delivery requirements rather than general geometric or legal design compliance; it does not establish the candidate's computability classification or an exact/bounded/UNKNOWN/official-review router.","source_ids":["S4"]},{"name":"CORENET X Model Checker","similarity":"This is an authority-backed checker integrated with digital regulatory submissions and explicitly limited to model quality and selected regulatory requirements.","remaining_difference":"Its public documentation describes a supplementary, scoped checker but does not document formal termination proofs, undecidability analysis, proof-gated fragments, or the candidate's explicit multi-guarantee routing contract.","source_ids":["S5","S6"]}],"distinctive_claim_remaining":"The remaining testable distinction is not UNKNOWN or human fallback alone, which S3 already implements. It is a proof-governed boundary for an authority's actual accepted plan-rule contract: formally classify its expressiveness, enforce a total/exact sublanguage at ingestion, attach machine-checkable termination and correctness obligations, and use those proof and bound labels to control routing without permitting Boolean label leakage.","confidence":"HIGH"},"implementation_evidence":{"support":"MODERATE","rationale":"Most components are demonstrably implementable: D-COM implemented tri-state routing, alternate-service fallback, human escalation, audit records, and manual comparison (S3); IDS demonstrates enforceable grammar restrictions and conformance tests (S4); and Lean documents practical structural and well-founded termination obligations (S2). What remains unsupported is the crucial domain bridge—whether a real jurisdiction's plan-rule language is already total or can encode unrestricted computation—and the legal/spatial fidelity of any proof model.","source_ids":["S2","S3","S4"]},"scores":{"meaningful_impact":{"score":4,"rationale":"Compliance decisions protect life, property, and public interest, and official evidence identifies inconsistent oversight and capacity constraints (S7). Preventing an unsupported automated clearance could be important, although the frequency of the exact Boolean/timeout failure is unknown.","source_ids":["S7"]},"stakeholder_pull":{"score":3,"rationale":"Mandatory digital submissions, multi-agency pilots, and official digital-backbone recommendations show substantial pull for digitized, accountable review (S6, S7). No source shows demand for computability proofs or dissatisfaction specifically caused by totality claims.","source_ids":["S6","S7"]},"incremental_advantage":{"score":2,"rationale":"Tri-state outcomes, alternate-service routing, human fallback, manual oversight, and audit trails already appear together in close prior art (S3). The incremental advantage is narrowed to proof-enforced language boundaries and guarantee-aware routing, whose operational benefit is untested.","source_ids":["S2","S3","S4"]},"distinctiveness_plausibility":{"score":2,"rationale":"The search found substantial collision with the proposed operational package (S3) and adjacent standards for enforceable checker scope (S4). A distinctive formal-computability layer may remain, but no domain implementation or exhaustive prior-art search establishes it.","source_ids":["S3","S4"]},"technical_implementability":{"score":3,"rationale":"A fixed fragment, conformance suite, explicit termination measure, router, and audit record are individually feasible (S2, S3, S4). The decisive uncertainties are semantic fidelity, useful coverage, and whether the real accepted language supports either a uniform totality proof or a faithful impossibility construction.","source_ids":["S2","S3","S4"]},"adoption_authority_feasibility":{"score":4,"rationale":"Government authorities already operate digital building-approval systems and scoped model checkers, and official guidance assigns compliance decisions to competent building-control bodies using professional judgment (S5, S6, S7). Shadow-only use with retained official authority is institutionally plausible.","source_ids":["S5","S6","S7"]},"evidence_readiness":{"score":4,"rationale":"Existing conformance suites, digitized rules, prototype architectures, and manual-comparison methods provide reusable study scaffolding (S3, S4). A frozen jurisdictional grammar and representative corpus still must be secured.","source_ids":["S3","S4"]},"safety_net_benefit":{"score":4,"rationale":"Preserving unknown results and escalating them to human decision is an implemented mechanism directly aligned with avoiding unsupported Boolean verdicts (S3). The benefit is credible but not distinctive, and model-to-reality errors remain outside a formal checker.","source_ids":["S3","S7"]},"scalability":{"score":2,"rationale":"The closest prototype required 390 labor-hours to digitize three documents, and stakeholders identified data standardization, accuracy, BIM adoption, and role concerns (S3). Cross-jurisdiction rule interpretation, proof maintenance, and reviewer capacity are likely recurring bottlenecks despite official progress on shared digital infrastructure (S6, S7).","source_ids":["S3","S6","S7"]}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"50K_TO_250K","scope":"A shadow-only differentiation study for one frozen egress-rule fragment and fixed archived/synthetic IFC corpus: formal semantics, parser enforcement, termination or counterexample work, D-COM-style tri-state baseline, proof-gated router, boundary-case injection, professional review, secure data handling, evaluation, coordination, and a legal-authority review.","confidence":"MODERATE","assumptions":["The jurisdiction supplies an existing encoded subset or permits a narrowly scoped encoding; S3's observed rule-digitization effort indicates that formalization is not negligible.","The team includes formal-methods, software/testing, architecture or code, official-review, evaluation, and coordination labor.","Standard workstations and modest secure compute are sufficient; no live permitting integration or automated decision authority is included.","The range incorporates loaded labor and overhead rather than multiplying published wages directly; S8 provides only a wage anchor, not a project price."],"source_ids":["S3","S8"]},"initial_deployment_startup":{"band_2026_usd":"250K_TO_1M","scope":"Controlled shadow-production integration for one authority and bounded fragments, including rule compiler and proof checks, permit-platform interfaces, access controls, audit records, monitoring, regression corpus, cybersecurity and privacy review, procurement and legal coordination, staff training, accessibility review, software and infrastructure.","confidence":"LOW","assumptions":["Existing permitting and identity systems expose usable interfaces.","No permit approval, denial, enforcement, or timeout reinterpretation is automated.","The scope is materially smaller than a full digital-permitting replacement but requires several professional roles and maintained controls, consistent with S3 and S7."],"source_ids":["S3","S7","S8"]},"operational_launch":{"band_2026_usd":"1M_TO_5M","scope":"Governed launch for one authority across a useful portfolio of bounded rules, with production integration, formally maintained semantics, data and model-quality controls, applicant disclosures, human escalation capacity, acceptance and safety evaluation, legal and procurement compliance, training, incident response, cybersecurity, records management, software licensing or development, and equipment or cloud operations.","confidence":"LOW","assumptions":["Useful-fragment coverage and manageable UNKNOWN volume have first been demonstrated.","Qualified officials retain every consequential decision.","The band excludes replacement of the whole permitting platform and excludes multi-jurisdiction rollout.","Costs reflect the coordination and data-standardization burdens documented in S3 and the institutional digital-infrastructure requirements in S6 and S7."],"source_ids":["S3","S6","S7","S8"]},"annual_recurring":{"band_2026_usd":"250K_TO_1M","scope":"Annual operation for one authority: rule and proof maintenance, code-change analysis, regression and adversarial testing, platform and security operations, audits, records retention, reviewer support, incident handling, vendor coordination, licenses or cloud resources, accessibility and legal review, evaluation, and periodic boundary recertification.","confidence":"LOW","assumptions":["Deployment remains limited to one authority and bounded fragments.","Material rule or schema changes trigger revalidation rather than silent expansion.","At least several fractional or full-time technical and domain roles are required; S3's digitization effort and S7's governance requirements support this resource assumption.","Major platform replacement and multi-jurisdiction scaling are excluded."],"source_ids":["S3","S7","S8"]}},"verified_pipeline_gates":{"externally_supported_problem":{"status":"UNCERTAIN","reason":"The safety significance and practical limitations of automated compliance checking are supported, but the bounded search found no actual authority demanding or vendor promising a universally terminating exact Boolean checker over an extensible plan-rule language, nor documented timeout-to-verdict leakage.","source_ids":["S3","S5","S7"]},"externally_credible_adopter_or_authorizer":{"status":"YES","reason":"Municipal and national building-control bodies demonstrably operate digital submission and checking workflows; Singapore's URA/BCA system and England's building-control framework are credible authorizer analogues.","source_ids":["S5","S6","S7"]},"distinct_testable_incremental_claim":{"status":"YES","reason":"Against the close S3 baseline, the narrow incremental claim is testable: proof-enforced language admission and guarantee-aware routing should prevent unsupported Boolean outputs beyond what a conventional tri-state/manual-fallback router achieves.","source_ids":["S2","S3","S4"]},"bounded_next_evidence_step":{"status":"YES","reason":"A frozen rule subset, fixed corpus, two-router comparison, formal totality attempt, adversarial boundary cases, and explicit unsoundness and no-benefit falsifiers form a bounded non-live study.","source_ids":["S2","S3","S4"]},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"A shadow-only study can retain all official decisions, treat UNKNOWN and timeouts as non-verdicts, use archived or synthetic cases, and stop on any unsound clearance or label leakage; this aligns with existing manual-oversight practice and official responsibility.","source_ids":["S3","S7"]},"credible_cost_scope_and_range":{"status":"YES","reason":"The broad bands enumerate labor, data, compliance, coordination, equipment or compute, software, security, and evaluation. They are anchored by observed rule-digitization effort and official professional-wage data, while appropriately retaining low confidence beyond first evidence.","source_ids":["S3","S8"]}},"next_evidence_step":"Preregister a shadow-only comparison using one authority's frozen machine-readable egress subset and a fixed archived/synthetic IFC corpus. First attempt to prove that every parser-admitted rule and plan belongs to a total, exact fragment; in parallel, attempt a domain-faithful counterexample only for constructs actually admitted. Compare (A) the closest-prior-art baseline—a true/false/UNKNOWN router with human escalation—with (B) the same router gated by machine-enforced fragment membership, proof obligations, and explicit execution bounds. Measure unsound clearances, in-scope misrouting, UNKNOWN rate, label leakage, reviewer minutes, and useful-case coverage. Falsify the problem if the complete legally accepted class is enforceably total with a uniform correct terminating procedure; falsify the intervention if it produces any unsound clearance or label leakage, excludes safety- or accessibility-critical cases without safe review, or offers no safety/reviewer-load improvement over baseline A.","blocking_evidence":["No evidence yet establishes that an authority or vendor actually makes the candidate's universal total-Boolean promise or converts timeouts into legal verdicts.","The computational expressiveness of any jurisdiction's legally accepted plan-rule contract remains unclassified.","No domain-faithful answer-preserving reduction from an unrestricted computation problem has been demonstrated.","The useful coverage of a proof-enforced total fragment, especially for atypical and accessibility-critical designs, is unknown.","The incremental operational value over the already-published true/false/UNKNOWN plus human-escalation architecture is unknown.","Reviewer capacity under realistic UNKNOWN and false-alarm rates is unknown.","The bounded search did not exhaust patents, proprietary systems, non-English research, municipal procurements, or unpublished implementations."],"research_disposition":"PRIOR_ART_DIFFERENTIATION_STUDY","world_novelty_boundary":"This bounded open-book search found substantial collision with an implemented extensible compliance ecosystem that already preserves UNKNOWN, tries alternate services, escalates to humans, retains manual oversight, and records audits. The only plausible remaining distinction is formal, proof-enforced computability classification tied to the actual admitted plan-rule language and used to govern routing. No source found an implementation of that complete distinction, but this is not a world-novelty claim; patents, proprietary products, non-English literature, municipal procurement records, and unpublished systems were not exhaustively searched."}