{"schema_version":1,"research_id":"eoa_inverse_innovation_exp04_external_evaluation_20260802","source_assessment_id":"layer_decay_and_expiration_management__film_media_production:SENTINEL_MATCHED:v0","cell_id":"layer_decay_and_expiration_management__film_media_production","search_queries":["site:loc.gov recommended format specifications moving image preservation file formats","site:oscars.org digital dilemma 2 digital motion picture archival storage post production files","SMPTE media asset management version control archive post production standard","post production storage version sprawl media asset management archive workflow official","site:help.iconik.backlight.co archive restore storage tiers versions assets official","site:knowledge.autodesk.com shotgrid versions media dependencies archive official","site:avid.com MediaCentral archive restore media lifecycle official","site:docs.aws.amazon.com Amazon S3 lifecycle transition expiration restore testing object lock legal hold","iconik official storage archive restore asset versions documentation","Autodesk Flow Production Tracking official version tracking media files versions assets","Frame.io version stacking official help versions assets archive storage","CatDV official archive restore media asset management storage lifecycle","NDSA Levels of Digital Preservation 2019 official fixity storage metadata access","ISO 14721 OAIS reference model preservation information packages official summary","National Archives audiovisual digital preservation file lifecycle fixity metadata official","DPC digital preservation storage archive restore testing fixity official guidance","Sundance Institute archive independent filmmakers digital preservation guide assets archive post production official","site:sundance.org filmmakers archive digital files preservation workflow post production","site:filmindependent.org archive film project files preservation media management"],"sources":[{"source_id":"S1","title":"The Digital Dilemma 2: Perspectives from Independent Filmmakers, Documentarians and Nonprofit Audiovisual Archives","publisher":"Academy of Motion Picture Arts and Sciences, in partnership with the Library of Congress NDIIPP","url":"https://www.oscars.org/sites/oscars/files/digital_dilemma2.pdf?VersionId=JCls1A6vxdkKPWlEmg3veHkrmzhbiCSO","source_class":"PRIMARY_RESEARCH","publication_date":"2012","accessed_at":"2026-08-02","claims_supported":["Surveys, interviews, and archival case studies found that digital motion-picture materials require active management and that independent filmmakers and nonprofit archives were under-resourced.","Few surveyed independent filmmakers used appropriate digital-master preservation practices, while surveyed archives anticipated rapid growth in digital holdings.","The report calls for preservation to become part of planning and budgeting and describes substantial perpetual capital and operating expense."]},{"source_id":"S2","title":"Archives and Preserve Your Work","publisher":"Sundance Institute","url":"https://www.sundance.org/archives","source_class":"OFFICIAL_ORGANIZATION_DATA","publication_date":"Undated; page copyright 2025","accessed_at":"2026-08-02","claims_supported":["Sundance reports 2.5 million archival assets, including more than 500,000 digital-video files.","Sundance identifies loss risks from poor storage, changing ownership, closing labs, and neglect and operates a preservation fund and UCLA partnership.","Its guidance recommends consistent labeling, multiple geographically separated copies, six-month media checks, periodic migration, and coordination with distributors while filmmakers retain rights."]},{"source_id":"S3","title":"Recommended Formats Statement — Moving Image Works","publisher":"Library of Congress","url":"https://www.loc.gov/preservation/resources/rfs/moving.html","source_class":"GOVERNMENT_OR_REGULATOR","publication_date":"Living statement; accessed 2026-08-02","accessed_at":"2026-08-02","claims_supported":["The Library distinguishes final production versions, masters, tracks, metadata, and viewing proxies rather than treating every artifact as equivalent.","It prefers identifiable, interoperable preservation packages and directs depositors to consult an archive about master and pre-production materials."]},{"source_id":"S4","title":"SMPTE ST 2067 — Interoperable Master Format","publisher":"Society of Motion Picture and Television Engineers","url":"https://www.smpte.org/standards/st2067","source_class":"STANDARD","publication_date":"Living standards page; accessed 2026-08-02","accessed_at":"2026-08-02","claims_supported":["IMF is an established framework for storing and exchanging multiple versions of a finished audiovisual work using shared essence and composition playlists.","Application variants address archival masters, preservation after digital post-production, version derivation, and storage-capacity efficiency."]},{"source_id":"S5","title":"ISO 14721:2025 — Reference Model for an Open Archival Information System (OAIS)","publisher":"International Organization for Standardization","url":"https://www.iso.org/standard/87471.html","source_class":"STANDARD","publication_date":"2025-03","accessed_at":"2026-08-02","claims_supported":["OAIS already defines an organizational and technical framework spanning ingest, archival storage, data management, access, dissemination, migration, and preservation responsibility.","Long-term preservation expressly includes changing formats, software, media, and designated-community knowledge."]},{"source_id":"S6","title":"Restoring an Asset from MediaCentral | Archive Production","publisher":"Avid Technology","url":"https://mediacentral.avid.com/mccux/Content/MCUX_Users_Guide/Restoring_an_Asset_from_MC_Archive.htm","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"Undated product documentation","accessed_at":"2026-08-02","claims_supported":["Avid already supports searchable archived production assets, retained metadata and restrictions, explicit archive state, configured restore profiles, and monitored restore jobs.","The documented workflow can retain low-resolution media locally while high-resolution media is archived."]},{"source_id":"S7","title":"Archive and Restore","publisher":"Backlight iconik","url":"https://help.iconik.backlight.co/hc/en-us/articles/25303782583703-Archive-and-Restore","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"2024","accessed_at":"2026-08-02","claims_supported":["Iconik already implements role-controlled movement of original media to archive storage, explicit archive status, restoration, and a marker when the active-storage copy is deleted.","The product demonstrates that reversible tier transitions and differentiated permissions are technically routine in media-asset systems."]},{"source_id":"S8","title":"Object Lock Considerations — Amazon Simple Storage Service","publisher":"Amazon Web Services","url":"https://docs.aws.amazon.com/AmazonS3/latest/userguide/object-lock-managing.html","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"Living documentation; accessed 2026-08-02","accessed_at":"2026-08-02","claims_supported":["S3 exposes inventory fields for retention dates and legal-hold status and preserves locks across lifecycle storage-class transitions.","Locked versions cannot be deleted by lifecycle expiration; delete markers can hide an object without destroying its retained versions.","The documentation identifies permission, encryption-key, integrity-check, logging, and governance-bypass constraints relevant to safe implementation."]}],"problem_evidence":{"support":"STRONG","rationale":"The Academy's survey and case-study evidence directly documents unmanaged digital-motion-picture preservation, rapidly growing holdings, limited resources, and ongoing access risk. Sundance independently reports a live collection of 2.5 million assets and explicitly describes loss from poor storage and neglect. SMPTE's treatment of multiple finished versions corroborates version proliferation. Evidence that obsolete versions specifically cause frequent wrong deliveries, retrieval delay, or material active-tier expense in present-day post-production facilities was not found, so those narrower consequences remain unmeasured.","source_ids":["S1","S2","S4"]},"stakeholder_evidence":{"support":"STRONG","rationale":"Sundance Institute is an identifiable archive operator and funder: it operates a preservation fund, maintains the Sundance Collection with UCLA, publishes preservation guidance, and invites filmmakers to deposit work. The Academy and Library of Congress jointly commissioned primary research calling for coordinated preservation planning and budgeting. This establishes credible institutional pull for preservation and lifecycle governance, but no source commits any organization to this exact dependency-aware pilot.","source_ids":["S1","S2","S3"]},"prior_art":{"proximity":"SUBSTANTIAL_COLLISION","closest_analogues":[{"name":"Avid MediaCentral Archive Production","similarity":"Provides searchable archive state, preserved metadata and restrictions, explicit restore profiles, and monitored restoration inside a production-management workflow.","remaining_difference":"The opened documentation does not show a cross-tool dependency graph, class-specific expiration rules, owner/legal exception workflow, reversible quarantine, or randomized recurring restore drills.","source_ids":["S6"]},{"name":"Backlight iconik Archive and Restore","similarity":"Provides permissioned archive and restore, movement between storage locations, visible lifecycle state, and a deleted-on-primary marker while an archived copy remains restorable.","remaining_difference":"The opened documentation does not establish semantic staleness scoring, dependency-safe disposition across edit/VFX/sound project structures, legal-hold adjudication, or comparative outcome measurement.","source_ids":["S7"]},{"name":"Amazon S3 Lifecycle, Object Lock, Inventory, and Delete Markers","similarity":"Implements retention metadata, legal holds, storage transitions, inventory, permissions, integrity checks, and reversible logical deletion primitives at scale.","remaining_difference":"It is a substrate-level object-storage mechanism without film-production identity, creative authority, proprietary-project dependency analysis, or authoritative-version semantics.","source_ids":["S8"]},{"name":"SMPTE IMF plus OAIS archival practice","similarity":"IMF composes multiple finished versions from shared essence, while OAIS covers ingest, storage, management, access, dissemination, and migration under organizational responsibility.","remaining_difference":"These standards do not themselves govern the lifecycle of heterogeneous work-in-progress edit files, VFX renders, proxies, stems, exports, and proprietary project dependencies or specify the proposed 30-day quarantine experiment.","source_ids":["S4","S5"]}],"distinctive_claim_remaining":"On one completed production, a governance overlay that resolves identity and observable dependencies across edit versions, VFX renders, proxies, sound stems, exports, and project files, then applies owner/legal-gated archive or 30-day quarantine transitions with persistent supersession markers and sampled end-to-end restores, will outperform a one-time manual archive sweep on median authoritative-artifact retrieval time, wrong-current selections, and active-storage bytes without increasing total review labor by more than 25% and with zero missed live dependencies or failed integrity/relink restores. This integrated outcome claim remains falsifiable; its individual mechanisms are substantially established.","confidence":"HIGH"},"implementation_evidence":{"support":"MODERATE","rationale":"SMPTE and Library of Congress specifications establish interoperable masters, composition metadata, identifiers, and distinct preservation/access derivatives. Avid and iconik demonstrate operational archive states, permissions, search, and restoration; AWS demonstrates lifecycle transitions, retention/hold metadata, inventory, integrity checks, and delete markers; OAIS establishes the broader archival workflow. The unresolved technical issue is dependency completeness across proprietary NLE/VFX/audio formats, plug-ins, external collaborators, and tacit practices. Legal and workflow feasibility requires production-owner, distributor, rights, consent, and hold review; the sources establish governance primitives but not a universal private-production retention schedule. A simulation and reversible quarantine are feasible, while automatic hard deletion is not currently justified.","source_ids":["S2","S3","S4","S5","S6","S7","S8"]},"scores":{"meaningful_impact":{"score":4,"rationale":"Loss and inaccessible digital work threaten cultural, creative, and revenue value, and the affected holdings can be very large. The candidate-specific frequency of wrong-version use and retrieval delay is not quantified.","source_ids":["S1","S2"]},"stakeholder_pull":{"score":4,"rationale":"Sundance operates preservation services and a fund, while the Academy and Library of Congress explicitly called for planning, budgeting, collaboration, and preservation infrastructure. Exact pilot commitment is absent.","source_ids":["S1","S2"]},"incremental_advantage":{"score":2,"rationale":"Archive/restore, role controls, markers, lifecycle transitions, multiple-version composition, and archival process models already exist. Advantage depends on proving that their film-specific orchestration beats a manual sweep.","source_ids":["S4","S5","S6","S7","S8"]},"distinctiveness_plausibility":{"score":2,"rationale":"The remaining distinction is an integrated dependency- and authority-aware operating protocol across heterogeneous post-production artifacts, not a clearly new component. Substantial prior-art overlap constrains distinctiveness.","source_ids":["S4","S5","S6","S7","S8"]},"technical_implementability":{"score":4,"rationale":"Most state, permission, inventory, tiering, marker, and restore primitives are commercially implemented. Hidden dependencies and legacy software/codec context prevent a maximum score.","source_ids":["S3","S4","S5","S6","S7","S8"]},"adoption_authority_feasibility":{"score":3,"rationale":"A production owner can authorize a completed-project pilot, and Sundance demonstrates coordination with filmmakers, distributors, and an archive while rights remain with the depositor. Actual partner authority and contractual schedules remain unverified.","source_ids":["S2","S8"]},"evidence_readiness":{"score":3,"rationale":"The problem, standards, and enabling products are documented, and the proposed metrics can be collected on a bounded sample. Contemporary prevalence, baseline error rates, labor costs, and comparative outcomes require proprietary workflow data.","source_ids":["S1","S2","S6","S7"]},"safety_net_benefit":{"score":4,"rationale":"Quarantine, archive copies, restore testing, role restrictions, legal holds, integrity checks, and delete markers materially reduce irreversible-loss risk. They cannot reveal dependencies that were never inventoried and can prolong sensitive-data retention.","source_ids":["S2","S6","S7","S8"]},"scalability":{"score":3,"rationale":"Inventory and lifecycle primitives scale technically, and Sundance reports millions of assets, but semantic identity resolution and human authority review may scale poorly across proprietary and externally held project components.","source_ids":["S2","S5","S8"]}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"10K_TO_50K","scope":"One 30-day, non-destructive comparative pilot covering 200 artifacts in two classes on one completed production: inventory, labels, dependency interviews/checks, simulated tiering, owner-approved quarantine, restores, and metric collection.","confidence":"LOW","assumptions":["Existing production storage and archive capacity are used.","Approximately 100-250 blended specialist hours are required across an assistant editor, archive/IT specialist, supervisor, and limited legal/rights review.","No preservation master is moved or altered, and no custom parser for a proprietary project format is built.","This is a resource-equivalent estimate, not a vendor quote."],"source_ids":["S1","S2","S6","S7"]},"initial_deployment_startup":{"band_2026_usd":"50K_TO_250K","scope":"Configure a lifecycle-state schema, policy matrix, role model, inventory connectors, audit log, dashboards, quarantine storage, restore procedure, and staff training for one post-production organization and several completed projects.","confidence":"LOW","assumptions":["Existing MAM/PAM and storage services expose usable APIs and metadata.","Approximately 500-1,500 engineering, archival, workflow-design, security, and training hours are required.","Storage hardware replacement and mass media migration are excluded.","No reliable public 2026 implementation quote was found."],"source_ids":["S4","S5","S6","S7","S8"]},"operational_launch":{"band_2026_usd":"250K_TO_1M","scope":"Production launch across multiple teams and 1-5 million existing assets, including backfill, identity reconciliation, connector hardening, rights/hold mapping, parallel validation, governance rollout, and incident/rollback preparation.","confidence":"LOW","assumptions":["Scale is comparable to Sundance's reported multi-million-asset collection.","Backfill requires both automation and substantial human exception review.","The organization already owns primary storage and production applications.","Large physical-media migration, enterprise license replacement, and destructive disposition are excluded."],"source_ids":["S1","S2","S5","S8"]},"annual_recurring":{"band_2026_usd":"50K_TO_250K","scope":"For a mid-sized organization: policy revalidation, exception and hold administration, restore drills, integrity monitoring, archive retrieval, metadata remediation, user support, and incremental storage/transfer costs.","confidence":"LOW","assumptions":["Approximately 0.5-1.5 full-time-equivalent staff plus storage and retrieval charges are required.","The archive is hundreds of terabytes rather than a major-studio multi-petabyte estate.","Sundance-style periodic checks and migration remain ongoing obligations.","The Academy supports the existence of substantial perpetual operating expense but its numerical evidence is historical, so it was not treated as a 2026 price benchmark."],"source_ids":["S1","S2","S5","S8"]}},"verified_pipeline_gates":{"externally_supported_problem":{"status":"YES","reason":"Academy survey/case-study evidence and Sundance's live collection establish ongoing digital preservation, access, scale, and resource problems, although current wrong-version and retrieval-delay prevalence remains unmeasured.","source_ids":["S1","S2","S4"]},"externally_credible_adopter_or_authorizer":{"status":"YES","reason":"Sundance Institute is an identifiable archive operator and preservation funder with a UCLA partnership, millions of assets, published need, and a contact path. This establishes a credible adopter/funder class, not a commitment to this candidate.","source_ids":["S2"]},"distinct_testable_incremental_claim":{"status":"YES","reason":"The remaining claim specifies an integrated intervention, a manual-sweep comparator, quantitative retrieval/error/storage/labor outcomes, and dependency/restore falsifiers.","source_ids":["S4","S6","S7","S8"]},"bounded_next_evidence_step":{"status":"YES","reason":"A 200-artifact, two-class, one-production, 30-day reversible pilot has bounded scope, comparators, stop conditions, and measurable outputs.","source_ids":["S2","S6","S7","S8"]},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The next step can be limited to a completed non-litigated production, simulation, and owner-approved reversible quarantine. Hard deletion, preservation-master alteration, disputed assets, unresolved holds, and failed dependency or restore checks remain prohibited.","source_ids":["S2","S7","S8"]},"credible_cost_scope_and_range":{"status":"UNCERTAIN","reason":"Resource scopes and broad bands can be bounded, but no current quotes, loaded labor rates, asset-volume profile, application inventory, or storage/retrieval bill for a willing pilot site were found. The bands are planning estimates rather than validated 2026 prices.","source_ids":["S1","S2","S6","S7","S8"]}},"next_evidence_step":"Recruit one production owner or post-production supervisor for a completed, non-litigated project. Stratify 200 artifacts across two classes, preferably edit/project files and one derivative-heavy class such as VFX renders or proxies. Randomly assign comparable folders or artifact families to (A) the existing one-time manual sweep and (B) the proposed lifecycle workflow. Before intervention, record active bytes, median time for blinded staff to find the authoritative artifact, wrong-current selections, review labor, known inbound references, holds, and baseline restore results. In arm B, inventory and identity-resolve artifacts, label lifecycle state and authority, simulate tiering, and quarantine only owner-approved duplicates or superseded derivatives for 30 days while retaining checksums, audit entries, supersession markers, and an archive copy. Run blinded retrieval tasks at days 0 and 30 and restore every quarantined artifact plus an age/format-stratified archived sample. Falsify the intervention if it fails to improve either median retrieval time or wrong-current selections, reduces active bytes by less than 10%, raises review labor more than 25%, misses any live dependency, produces any integrity/relink failure, exceeds the agreed restore-time target, or encounters disputed authority or a hold. Restore quarantined artifacts and stop immediately on any safety falsifier.","blocking_evidence":["No contemporary representative audit quantifies how much post-production storage is obsolete, how often obsolete artifacts appear current, or how much retrieval delay is attributable to lifecycle failure rather than naming/search defects.","No named production owner or post-production facility has committed proprietary asset and workflow data to the pilot.","Dependency-graph recall for proprietary NLE, VFX, audio, plug-in, external-collaborator, and dynamically resolved references is unknown.","No comparative live evidence shows improvement over the one-time manual-sweep rival.","No jurisdiction- and contract-specific retention, consent, privacy, distributor, insurer, or litigation-hold schedule has been reviewed for a pilot production.","The 2026 cost bands lack current labor, integration, licensing, storage, retrieval, and migration quotes from an actual site.","World novelty, patentability, freedom to operate, market size, and realized impact were not measured."],"research_disposition":"PARTNERED_RESEARCH_PROGRAM","world_novelty_boundary":"The eight opened sources establish substantial collision with archival standards, multiple-version mastering, production archive/restore products, and object-lifecycle governance primitives. They do not establish that the exact cross-artifact composition has or has not been implemented elsewhere. This evaluation therefore makes no world-novelty, patentability, freedom-to-operate, market-size, or realized-impact determination; novelty is bounded to the untested contrastive workflow-and-outcome claim described here.","arm":"SENTINEL","candidate_version":0,"controller_recommendation":{"action":"STOP_EMPIRICAL_RESEARCH_NEEDED","repairable":false,"material_progress_observed":true,"progress_targets":["Obtain written participation and lifecycle authority from one production owner or post-production supervisor for a completed, non-litigated project.","Measure baseline prevalence, authoritative-version errors, retrieval time, active-storage bytes, review labor, dependency coverage, and restore success on the bounded sample.","Run the randomized or matched manual-sweep comparator versus the proposed lifecycle workflow for 30 days.","Demonstrate zero missed dependencies, zero integrity/relink failures, and compliance with all holds and owner approvals.","Validate the four cost bands with site-specific loaded labor, licensing, integration, storage, retrieval, and migration quotes.","Document whether any current MAM/PAM already supplies the complete dependency-, authority-, quarantine-, marker-, and restore-drill workflow, narrowing the remaining incremental claim."],"reason":"Bounded web research verified the problem class, credible stakeholders, enabling technology, standards, and substantial prior-art overlap. The remaining decision turns on proprietary dependency coverage, site authority, comparative workflow outcomes, restore behavior, and actual cost—all requiring a partner's data and live testing. Under the requested controller rule this is a stop for empirical research, and stop recommendations are marked non-repairable."}}