{"schema_version":1,"research_id":"eoa_inverse_innovation_exp05_external_evaluation_20260803","source_assessment_id":"computability_boundary_mapping__human_computer_interaction:P3:v0","cell_id":"computability_boundary_mapping__human_computer_interaction","search_queries":["site:justice.gov web mobile app accessibility rule barriers people with disabilities state local governments 2024","site:section508.gov accessibility modernization legacy systems playbook ICT federal agencies","site:w3.org/TR/WCAG22 WCAG 2.2 Recommendation","Wizard of Oz prototyping hidden human operator limitations automation evidence HCI paper","Turing 1936 computable numbers halting problem original paper PDF","official model checker equivalence checking finite state systems documentation bisimulation","automatic user interface migration accessibility transformation research paper legacy applications","BLS software developers median pay 2025 UX accessibility specialist wages","site:mcrl2.org equivalence checking bisimulation finite state official documentation","site:bls.gov/ooh computer information technology software developers quality assurance analysts testers median pay 2025","site:bls.gov/oes/current software developers annual mean wage 2025","site:acm.org \"Mapping the Wizards' Path\" 194 papers","mCRL2 toolset behavioural equivalence official user manual bisimulation","site:cadp.inria.fr equivalence checking finite state systems official","site:spinroot.com finite state verification model checking official","finite automata equivalence checking algorithm official university notes","site:w3.org/TR model based user interfaces task models transformations abstract user interface recommendation","W3C model based UI legacy application migration transformation accessibility","automatic generation personalized user interfaces SUPPLE accessibility paper ACM","user interface migration task model finite state machine legacy application research","\"Mapping the Wizards' Path\" systematic review Wizard of Oz HCI PDF","\"A wrapping approach for migrating legacy system interactive functionalities\" PDF","\"On Computable Numbers, with an Application to the Entscheidungsproblem\" PDF university","Wizard of Oz HCI systematic review operator inconsistency cognitive load primary paper"],"sources":[{"source_id":"S1","title":"Technology Accessibility Playbook—Introduction","publisher":"Section508.gov, U.S. General Services Administration","url":"https://www.section508.gov/manage/playbooks/technology-accessibility-playbook/","source_class":"OFFICIAL_GUIDANCE","publication_date":"2025-09","accessed_at":"2026-08-03","claims_supported":["Federal agencies report that too many public-facing and employee digital services remain inaccessible.","Section 508 creates an institutional obligation to develop, procure, maintain, and use accessible ICT.","Federal CIO, CAO, and Section 508 offices are identifiable authorizers and adopters.","The playbook expressly calls for governance, resourcing, testing, validation, accessibility checkpoints, and integration with modernization."]},{"source_id":"S2","title":"Justice Department to Publish Final Rule to Strengthen Web and Mobile App Access for People with Disabilities","publisher":"U.S. Department of Justice","url":"https://www.justice.gov/archives/opa/pr/justice-department-publish-final-rule-strengthen-web-and-mobile-app-access-people","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"2024-04-08","accessed_at":"2026-08-03","claims_supported":["Inaccessible websites and mobile applications can make critical government services difficult or impossible for disabled people to use.","The ADA Title II rule establishes accessibility obligations and technical standards for state and local government web content and mobile applications.","Accessibility failures can exclude people from services including healthcare, courts, voting, education, taxes, and transit."]},{"source_id":"S3","title":"Introduction to Model-Based User Interfaces","publisher":"World Wide Web Consortium","url":"https://www.w3.org/TR/mbui-intro/","source_class":"OFFICIAL_GUIDANCE","publication_date":"2014-01-07","accessed_at":"2026-08-03","claims_supported":["Model-based UI engineering already represents tasks, domains, abstract interfaces, concrete interfaces, final interfaces, users, platforms, and environments.","Existing practice includes abstraction, reverse engineering, translation, transformation, code generation, runtime adaptation, and migratory interfaces.","The document emphasizes consistency between models and transformed implementations but is a Working Group Note rather than a W3C-endorsed Recommendation.","This is close prior art for task-semantic contracts and interface transformation, but it does not supply the proposed Wizard-to-Build computability gate."]},{"source_id":"S4","title":"Mapping the Wizards' Path: A Systematic Review of Wizard-of-Oz in HCI","publisher":"Association for Computing Machinery","url":"https://doi.org/10.1145/3772318.3791174","source_class":"PRIMARY_RESEARCH","publication_date":"2026-04-13","accessed_at":"2026-08-03","claims_supported":["Wizard-of-Oz is established HCI practice in which apparently autonomous functions are executed by hidden human operators.","A systematic review of 194 SIGCHI papers found accessibility and aging among the application domains and identified technical-limit, flexibility, cost, safety, and ethics motivations.","Reported concerns include wizard variability and bias, ecological validity, deception and ethics, and transparency and reproducibility.","The review recommends standardizing and documenting wizard behavior, supporting the proposal's provenance premise while not establishing the specific migration overclaim's prevalence."]},{"source_id":"S5","title":"On Computable Numbers, with an Application to the Entscheidungsproblem","publisher":"Proceedings of the London Mathematical Society / Wiley","url":"https://londmathsoc.onlinelibrary.wiley.com/doi/10.1112/plms/s2-42.1.230","source_class":"PRIMARY_RESEARCH","publication_date":"1937","accessed_at":"2026-08-03","claims_supported":["Turing formalized machine computation, universal simulation, and machine-unsolvable decision questions.","The result supplies the foundational computability boundary for a candidate-specific reduction.","It does not by itself prove that the proposal's chosen task-semantic observation relation has the required reduction; that construction remains an explicit proof obligation."]},{"source_id":"S6","title":"CADP Home Page and Toolset Description","publisher":"CONVECS, Inria","url":"https://cadp.inria.fr/","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"2026-07-14","accessed_at":"2026-08-03","claims_supported":["A maintained toolset already accepts automata and networks of finite-state machines.","Existing tools perform equivalence checking under bisimulation relations, model checking, simulation, on-the-fly verification, symbolic verification, and compositional minimization.","Exact finite-model comparison is therefore technically established, although adapting it to an HCI task-semantic contract and controlling state growth remain project work."]},{"source_id":"S7","title":"A Wrapping Approach for Migrating Legacy System Interactive Functionalities to Service Oriented Architectures","publisher":"Journal of Systems and Software, Elsevier","url":"https://www.sciencedirect.com/science/article/pii/S0164121207001380","source_class":"PRIMARY_RESEARCH","publication_date":"2008-04","accessed_at":"2026-08-03","claims_supported":["Black-box modernization of interactive legacy applications is established research practice.","The approach models legacy interaction behavior as finite-state machines and uses a wrapper to act on behalf of users.","Two migration case studies provide implementation evidence for bounded, model-based legacy-interface migration.","The work addresses selected wrapped functionality, not universal accessible replacement or unrestricted equivalence certification."]},{"source_id":"S8","title":"Software Developers, Quality Assurance Analysts, and Testers","publisher":"U.S. Bureau of Labor Statistics","url":"https://www.bls.gov/ooh/computer-and-information-technology/software-developers.htm","source_class":"OFFICIAL_ORGANIZATION_DATA","publication_date":"2025-08-28","accessed_at":"2026-08-03","claims_supported":["The May 2024 median annual wage was $133,080 for software developers and $102,610 for software QA analysts and testers.","The listed work includes requirements analysis, modeling, testing, risk identification, documentation, usability assessment, and stakeholder feedback.","These wage data provide a public labor-cost anchor for resource-equivalent estimates, but not a vendor quote or organization-specific budget."]}],"problem_evidence":{"support":"MODERATE","rationale":"The accessibility problem and its consequences are directly documented by federal sources, while current HCI research confirms that Wizard-of-Oz prototypes hide human execution and create transparency, variability, and reproducibility concerns. Legacy interactive-system migration is also an established activity. However, no located source documents how often organizations specifically infer a total converter and unrestricted equivalence certifier from successful Wizard-of-Oz sessions. The exact proposed failure mode is therefore plausible and consequential but not prevalence-verified.","source_ids":["S1","S2","S4","S7"]},"stakeholder_evidence":{"support":"MODERATE","rationale":"Federal CIOs, CAOs, Section 508 program offices, accessibility subject-matter experts, and IT governance bodies are identifiable adopters or authorizers. GSA expressly calls for accessibility governance, resourcing, validation, modernization integration, and authoritative accessibility checkpoints. DOJ supplies a regulatory incentive. No source expresses demand for this particular realizability gate, provenance alphabet, or computability-boundary workflow, so intervention-specific pull remains unverified.","source_ids":["S1","S2"]},"prior_art":{"proximity":"ADJACENT_PRIOR_ART","closest_analogues":[{"name":"Rigorous Wizard-of-Oz reporting and control practices","similarity":"Documents hidden operator behavior, variability, transparency, reproducibility, deception, and the need to standardize wizard operations.","remaining_difference":"Does not distinguish computability from prototyped experience, define a migration-equivalence contract, or route unrestricted cases to explicit unknown.","source_ids":["S4"]},{"name":"W3C model-based user-interface engineering and migratory interfaces","similarity":"Uses task and domain models, abstraction, reverse engineering, transformations, context-of-use models, code generation, and migratory interfaces.","remaining_difference":"Does not claim a checked undecidability boundary, provenance ledger, enforceable finite exact profile, or safety-preserving result alphabet; the source is also a non-endorsed Working Group Note.","source_ids":["S3"]},{"name":"Finite-state wrapper migration for interactive legacy systems","similarity":"Migrates selected legacy interactive functionality using a black-box wrapper whose behavior is represented by finite-state machines.","remaining_difference":"Wraps bounded use cases rather than producing an accessible replacement and does not certify equivalence of arbitrary executable interfaces or govern Wizard-of-Oz evidence.","source_ids":["S7"]},{"name":"CADP finite-state equivalence and model-checking toolset","similarity":"Provides mature finite-state modeling, equivalence checking, simulation, counterexample-related verification techniques, and scalable algorithmic variants.","remaining_difference":"Is a general formal-methods toolbox, not an HCI migration workflow; semantic fidelity, accessibility review, provenance, scope membership, and sponsor-facing labels must be supplied externally.","source_ids":["S6"]},{"name":"Turing-style computability boundary","similarity":"Establishes that some universal machine-behavior decision questions have no total mechanical solution.","remaining_difference":"Does not establish the proposal's specific task-semantic equivalence reduction or its observation-contract fidelity; those require a separate checked proof.","source_ids":["S5"]}],"distinctive_claim_remaining":"For a Wizard-of-Oz-based accessible legacy-interface migration, a pre-build gate that jointly records response provenance, fixes a nonvisual task-semantic observation contract, enforces a finite exact-certification profile, and labels bounded unrestricted search as WITNESSED_DIVERGENCE or UNKNOWN_EQUIVALENCE will reduce unsupported equivalence claims without discarding valid accessibility findings. This is falsifiable by fixture-level label accuracy, provenance retention, replayability, reviewer agreement, and whether sponsor decisions differ from scripted-task or screenshot baselines.","confidence":"MODERATE"},"implementation_evidence":{"support":"MODERATE","rationale":"All principal technical ingredients exist separately: task/domain UI models and transformations, finite-state legacy wrappers, maintained finite-state equivalence tools, and foundational undecidability methods. An offline fixture audit is within ordinary sponsor and accessibility-governance authority. The unverified work is integration: defining a semantically faithful observation contract, enforcing profile membership, constructing and independently checking the exact reduction, preserving provenance through sponsor-facing workflow, handling environmental and assistive-technology observations, and controlling state explosion. No live implementation of the integrated gate was found.","source_ids":["S1","S3","S4","S5","S6","S7"]},"scores":{"meaningful_impact":{"score":4,"rationale":"False assurance could exclude disabled users from important services or fund an impossible unrestricted certification requirement. The exact failure mode's prevalence is unknown, limiting the maximum score.","source_ids":["S1","S2"]},"stakeholder_pull":{"score":3,"rationale":"Federal accessibility governance has named authorities, legal incentives, and an expressed need for validation and modernization controls, but no adopter has requested this particular gate.","source_ids":["S1","S2"]},"incremental_advantage":{"score":3,"rationale":"The proposal usefully connects practices normally kept separate and provides honest UNKNOWN and HUMAN_DEPENDENT states. Each major component already has strong adjacent prior art, and comparative workflow benefit is untested.","source_ids":["S3","S4","S6","S7"]},"distinctiveness_plausibility":{"score":3,"rationale":"No close integrated analogue was found among the eight sources, but the proposal is principally a governance-and-integration composition of established WoZ reporting, model-based UI migration, finite-state checking, and computability theory.","source_ids":["S3","S4","S5","S6","S7"]},"technical_implementability":{"score":4,"rationale":"A bounded fixture, provenance ledger, explicit status alphabet, finite-profile checker, and fair counterexample search are technically credible. Semantic fidelity, reduction correctness, and state-space growth are material but testable risks.","source_ids":["S3","S5","S6","S7"]},"adoption_authority_feasibility":{"score":4,"rationale":"CIO, CAO, Section 508, accessibility, product, and IT-governance roles can authorize an offline gate and constrain production migration. Cross-functional approval adds coordination cost but no apparent authority stop.","source_ids":["S1","S2"]},"evidence_readiness":{"score":3,"rationale":"The next experiment can be performed on nonproduction fixtures with existing methods, but decisive evidence requires construction and live execution of the integrated audit rather than more bounded web research.","source_ids":["S4","S6","S7"]},"safety_net_benefit":{"score":4,"rationale":"Explicit unknown, human-dependent, out-of-scope, and witnessed-divergence labels directly limit false-equivalence assurance while leaving accessibility evaluation intact. Mis-specified observation contracts could still create false confidence.","source_ids":["S1","S2","S4","S6"]},"scalability":{"score":3,"rationale":"The decision structure is reusable, but observation contracts and profile mappings are application-specific, while exact finite-state comparison may suffer state explosion and unrestricted search may return many unknowns.","source_ids":["S3","S6","S7"]}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"50K_TO_250K","scope":"Design and execute one offline fixture audit: freeze the semantic contract and profile, implement provenance and result labels, adapt an existing finite-state comparison tool, construct seeded fixtures, obtain independent formal and accessibility review, and report results.","confidence":"MODERATE","assumptions":["Approximately 0.6-1.2 labor-years across formal methods, accessibility/HCI, QA, and project governance over 3-5 months.","Uses existing model-checking infrastructure rather than building an equivalence engine from scratch.","Loaded labor and specialist contracting are estimated at roughly 1.5-2.2 times cash wages.","No production-system integration or participant recruitment is included."],"source_ids":["S1","S4","S6","S8"]},"initial_deployment_startup":{"band_2026_usd":"250K_TO_1M","scope":"Harden the audit into a reusable internal service: contract schema, profile-membership validator, checker adapters, provenance capture, replay tooling, CI integration, documentation, access controls, and reviewer training.","confidence":"LOW","assumptions":["Approximately 2-5 labor-years across software, formal verification, accessibility, QA, and governance.","Existing open or institutionally licensed formal tools are reused.","Covers one legacy-interface technology family and one organization.","Excludes full production migration and broad portfolio onboarding."],"source_ids":["S1","S3","S6","S8"]},"operational_launch":{"band_2026_usd":"1M_TO_5M","scope":"Launch across an initial migration program: integrate multiple legacy runtimes, validate assistive-technology observations, security-review tooling, train operators and reviewers, establish approval workflow, and audit several representative applications.","confidence":"LOW","assumptions":["Approximately 6-15 labor-years plus integration, security, accessibility-lab, and program-management overhead.","Legacy applications require bespoke adapters and semantic-contract workshops.","Includes governance and rollout but not rewriting the legacy applications themselves.","State-space mitigation and tool hardening are required for operational workloads."],"source_ids":["S1","S3","S6","S7","S8"]},"annual_recurring":{"band_2026_usd":"250K_TO_1M","scope":"Maintain profiles, adapters, checker versions, provenance records, reviewer capacity, periodic audits, incident handling, and reclassification when languages, task commitments, or assistive technologies change.","confidence":"LOW","assumptions":["Approximately 2-5 ongoing FTE-equivalents plus compute and accessibility-testing expenses.","Application onboarding remains selective rather than enterprise-universal.","Does not include the cost of implementing replacement interfaces or remediating discovered accessibility defects.","Annual independent assurance and staff training are included."],"source_ids":["S1","S6","S8"]}},"verified_pipeline_gates":{"externally_supported_problem":{"status":"UNCERTAIN","reason":"Accessibility barriers and Wizard-of-Oz transparency risks are externally supported, but the defining claim that teams generalize bounded wizard sessions into unrestricted automatic conversion and total equivalence certification was not directly documented.","source_ids":["S1","S2","S4","S7"]},"externally_credible_adopter_or_authorizer":{"status":"YES","reason":"Federal CIOs, CAOs, Section 508 program managers, accessibility subject-matter experts, and IT governance bodies are named authorities with expressed responsibility for accessibility, validation, resourcing, and modernization controls.","source_ids":["S1","S2"]},"distinct_testable_incremental_claim":{"status":"YES","reason":"The integrated gate can be compared with scripted-task and finite-checker baselines using seeded equivalence/divergence fixtures, provenance loss, unknown-label integrity, replayability, and sponsor interpretation.","source_ids":["S3","S4","S6","S7"]},"bounded_next_evidence_step":{"status":"YES","reason":"A frozen, nonproduction fixture family with predetermined comparators, outputs, pass conditions, and falsifiers can test the proposed evidence semantics without deploying a replacement.","source_ids":["S4","S6","S7"]},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The first step is offline, leaves production unchanged, assigns approval to existing sponsor and accessibility authorities, and explicitly prohibits migration or universal-equivalence claims. The main safety risk—an incomplete semantic contract—is itself a halt condition.","source_ids":["S1","S2"]},"credible_cost_scope_and_range":{"status":"YES","reason":"The four ranges have explicit labor, integration, rollout, and maintenance scopes and are anchored to public developer and QA wages. They remain low-confidence planning estimates rather than quotes.","source_ids":["S1","S6","S8"]}},"next_evidence_step":"With one willing accessibility program and one formal-methods partner, build and run a 12-pair nonproduction fixture study under a frozen task-semantic contract: three equivalent finite pairs with intentionally different presentations; four finite pairs with seeded persistent-outcome, error, focus, or assistive-technology-announcement differences; two unrestricted pairs with early or delayed replayable divergence; one pair with a nonterminating handler preceding a later finite divergence; and two pairs with no divergence inside the bound. Compare (A) scripted-task/screenshot review, (B) finite checking without provenance or an explicit unknown contract, and (C) the full gate. Pre-register exact-verdict accuracy, false-equivalence rate, divergence replay rate, starvation, provenance retention, UNKNOWN_EQUIVALENCE preservation in sponsor-facing output, reviewer time, and sponsor interpretation. Falsify the intervention if profile membership is bypassable, any seeded finite pair is misclassified, nontermination starves a later trace, a divergence cannot be replayed, human behavior is recorded as generated, bounded non-finding is displayed as equivalence, or accessibility reviewers identify a material task commitment outside the contract.","blocking_evidence":["No direct external evidence establishes the prevalence of the proposal's defining Wizard-of-Oz-to-universal-equivalence overclaim.","No identified organization has expressed demand or committed resources for this specific realizability gate.","The candidate-specific halting reduction and finite comparison proof have not been constructed and independently checked under a concrete observation contract.","No live result shows that profile membership is enforceable, seeded verdicts are correct, dovetailing avoids starvation, provenance survives workflow handoffs, or UNKNOWN is not consumed as equivalent.","It is unknown whether a finite migration profile covers enough valuable legacy behavior to justify adoption.","It is unknown whether accessibility reviewers and users judge the proposed task-semantic contract complete and faithful.","Cost ranges lack vendor quotes, organization-specific legacy inventories, integration measurements, and measured state-space workloads."],"research_disposition":"PARTNERED_RESEARCH_PROGRAM","world_novelty_boundary":"World novelty, patentability, freedom to operate, market size, and realized impact remain unmeasured. This assessment searched only eight direct public sources for problem evidence, stakeholders, adjacent practice, technical feasibility, and costs; it is neither an exhaustive literature search nor a patent search.","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_version":0,"controller_recommendation":{"action":"STOP_EMPIRICAL_RESEARCH_NEEDED","repairable":false,"material_progress_observed":true,"progress_targets":["Obtain one named accessibility-program sponsor and formal-review partner for the offline fixture study.","Document whether real migration requirements contain unrestricted conversion, termination, or equivalence language and how Wizard-of-Oz evidence is currently represented.","Construct and independently verify the candidate-specific reduction against the frozen task-semantic observation contract.","Demonstrate enforceable finite-profile membership and correct exact verdicts across all seeded finite fixtures.","Demonstrate fair divergence search despite a nonterminating handler and replay every reported divergence.","Measure provenance retention and confirm that bounded non-finding remains UNKNOWN_EQUIVALENCE in sponsor-facing decisions.","Have disabled-user or accessibility representatives review semantic-contract coverage independently of formal equivalence.","Replace labor-only cost assumptions with measured engineering time, state-space workload, adapter effort, and organization-specific rollout estimates."],"reason":"Web research materially supports accessibility importance, credible authorities, Wizard-of-Oz transparency concerns, model-based migration, and finite-state equivalence tooling. It does not establish prevalence of the exact overclaim or validate the integrated gate. The remaining decisive evidence requires access to a real workflow, a concrete formalization, independent proof review, and live execution of fixtures; under the controller rule this requires STOP_EMPIRICAL_RESEARCH_NEEDED, and every STOP is non-repairable."},"proposal_index":3}