{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp09_archetype_breadth150_20260804","research_id":"eoa_inverse_innovation_exp09_light_prior_art_20260804","cell_id":"context_keyed_representation_switching__chemistry_materials","search_lanes":{"direct_problem_and_intervention":{"queries":["near infrared spectroscopy epoxy cure monitoring degree of cure formulation calibration filler scattering","NIR cure monitoring epoxy formulation specific calibration chemometrics","different epoxy formulations NIR degree cure calibration model","filled epoxy NIR spectroscopy filler scattering cure calibration formulation"],"source_ids":["SRC1","SRC2"],"no_result_note":null},"synonyms_and_historical_terms":{"queries":["near infrared cure monitoring epoxy resin systems formulation dependent spectra","The determination of the degree of cure in epoxy paints infrared spectroscopy","near infrared attenuated total reflectance epoxy amine cure reactions","DRFT-NIR epoxy conversion heterogeneous paints"],"source_ids":["SRC1","SRC2"],"no_result_note":null},"products_practices_and_standards":{"queries":["ASTM E1655 multivariate infrared quantitative analysis calibration validation standard","NIR software product identification select calibration model product recipe chemometrics","Bruker OPUS process spectroscopy scenario multiple products calibration methods","site:bruker.com OPUS process spectroscopy scenario multiple products calibration methods"],"source_ids":["SRC3","SRC4"],"no_result_note":null},"component_combination":{"queries":["process analytical technology recipe selects calibration model spectrometer method version","process NIR analyzer multiple products calibration model automatic product selection","inline NIR spectrometer recipe management calibration model batch product","shared NIR measurement channel different products different calibration models external trigger"],"source_ids":["SRC2","SRC3","SRC4"],"no_result_note":null}},"sources":[{"source_id":"SRC1","title":"Usage of Near-Infrared Spectroscopy for Inline Monitoring the Degree of Curing in RTM Processes","publisher":"MDPI, Polymers","url":"https://www.mdpi.com/2073-4360/13/18/3145","source_type":"PRIMARY_RESEARCH","claims_supported":["Inline NIR measurements and PLS calibration models can estimate degree of cure for an epoxy resin during resin-transfer molding.","Calibration performance depends on reference-spectrum choice, spectral preprocessing, and validation against reference cure measurements.","Spectral baselines can change with physical effects such as particle-size and color changes, so the spectrum-to-cure relationship is not purely chemistry-independent."]},{"source_id":"SRC2","title":"The determination of the degree of cure in epoxy paints by infrared spectroscopy","publisher":"Elsevier, Polymer Testing (article hosted on ResearchGate)","url":"https://www.researchgate.net/publication/229284767_The_determination_of_the_degree_of_cure_in_epoxy_paints_by_infrared_spectroscopy","source_type":"PRIMARY_RESEARCH","claims_supported":["The study evaluated distinct heterogeneous epoxy/polyamine and epoxy/polyamide paint formulations using NIR cure measurements.","Pigments and other inorganic components caused radiation-scattering baselines, while formulation components could overlap the oxirane band.","The authors warned that mechanically applying the spectrometric method where a curing-agent band overlaps the cure-indicating band would create important systematic error."]},{"source_id":"SRC3","title":"ASTM E1655-17(2024): Standard Practices for Infrared Multivariate Quantitative Analysis","publisher":"ASTM International","url":"https://store.astm.org/standards/e1655","source_type":"OFFICIAL_STANDARD","claims_supported":["ASTM E1655 covers development and validation of multivariate NIR and MIR calibration models.","The standard states that calibration procedures define the valid measurement range and the types of materials for which estimates are meaningful.","It distinguishes calibration-time validation from ongoing validation of predictions and explicitly recognizes interferences and matrix effects."]},{"source_id":"SRC4","title":"CMET 3.5: Software for Inline Process Control","publisher":"Bruker Optics GmbH & Co. KG","url":"https://www.bruker.com/en/products-and-solutions/infrared-and-raman/opus-spectroscopy-software/process-reaction-monitoring/_jcr_content/root/sections/section/sectionpar/linklist/contentpar-1/calltoaction.download-asset.pdf/links/item0/CMET_Flyer_EN.pdf","source_type":"FIRST_PARTY_PRODUCT","claims_supported":["Bruker CMET supports batch processes with changing products on process FT-NIR analyzers.","Different products, and therefore different calibration models, can be assigned to the same measurement channel.","Product-specific calibration parameters, external DCS triggers, current-measurement status signals, and result logging or archiving are supported."]}],"problem_evidence":{"status":"PARTLY_SUPPORTED","finding":"The underlying problem is visible: primary studies show that inline NIR can estimate epoxy cure, but filler scattering, baseline behavior, curing-agent absorption, reference choice, and preprocessing affect the correspondence between spectra and conversion. ASTM likewise bounds calibrations to validated material types and ranges. SRC2 provides especially direct evidence that applying a cure-spectral method across an interfering formulation can cause systematic conversion error. The retained sources do not directly demonstrate the proposal's specific operational failure mode in which updating a pooled model on formulation B silently degrades later estimates for formulation A.","source_ids":["SRC1","SRC2","SRC3"]},"closest_prior_art":[{"name":"Bruker CMET product-specific calibration routing on a shared process FT-NIR channel","source_ids":["SRC4"],"overlap":"It already assigns multiple products and their different calibration models to a shared measurement channel, accepts external process-control triggers, reports current measurement status, and archives results.","remaining_difference":"The retained material does not disclose operator-confirmed recipe and batch-metadata agreement, mandatory abstention for missing or conflicting keys, immutable inactive-map checkpoints, cross-map update-interference regression, or archived re-entry replay before reactivation."},{"name":"Formulation-aware NIR cure measurement for heterogeneous epoxy/polyamine and epoxy/polyamide paints","source_ids":["SRC2"],"overlap":"It treats two epoxy formulations, observes scattering and formulation-component interference, and warns that mechanical application of the wrong spectral interpretation can create systematic cure error.","remaining_difference":"It is an analytical method study rather than a shared inline controller with explicit context routing, version isolation, switching guards, abstention, or re-entry testing."},{"name":"ASTM E1655 calibration-domain and validation practice","source_ids":["SRC3"],"overlap":"It establishes that multivariate IR calibrations have bounded material domains and require validation, providing the standards basis for formulation-specific qualification and held-out testing.","remaining_difference":"It does not prescribe a two-map recipe router, inactive-map freezing, active-map display, rollback checkpoint, or cross-map noninterference test."},{"name":"Inline epoxy-cure PLS calibration demonstrated in RTM","source_ids":["SRC1"],"overlap":"It demonstrates the proposed sensing substrate and reference-calibrated cure estimator, including sensitivity to preprocessing and reference-spectrum choices.","remaining_difference":"Only one resin system is modeled, with no alternating-formulation map selection or model-update isolation."}],"prior_art_disposition":"ADJACENT_PRIOR_ART","contrastive_claim_remaining":"On one shared inline spectrometer, an externally confirmed recipe key must resolve to exactly one visible, approved, versioned cure map; missing or conflicting keys must suppress estimation; updating formulation B must be unable to change formulation A's locked outputs; and archived A cases must reproduce before A is reactivated. Product-specific calibration selection alone is already disclosed by SRC4, so only this combined selection-isolation-abstention-re-entry assurance remains contrastive in the retained record.","contrastive_claim_falsifier":"The contrast would be falsified by a public system or established practice that already couples recipe or batch-record routing on one process spectrometer with approved immutable per-product calibration versions, conflict-triggered abstention, automatic cross-model noninterference checks, and re-entry replay. Independently, the proposed benefit would fail if a preregistered replay showed that one fixed global model meets the same per-formulation error and endpoint criteria and remains stable after formulation-specific updates.","gates":{"adequate_source_search":{"status":"PASS","rationale":"Four search lanes covered the direct epoxy-cure problem, older NIR/DRFT terminology, an official calibration standard, commercial process-spectroscopy practice, and combinations of shared-channel product routing and model management. Exactly four opened sources from four publishers were retained, including primary research, an official standard, and a first-party product source.","source_ids":["SRC1","SRC2","SRC3","SRC4"]},"supported_problem":{"status":"PASS","rationale":"The problem is partly supported: formulation components, scattering, baselines, references, and preprocessing can alter NIR cure interpretation, and direct research identifies systematic error from spectral interference. Cross-formulation degradation after model updating remains unverified and is appropriately assigned to the next test.","source_ids":["SRC1","SRC2","SRC3"]},"distinct_testable_claim":{"status":"PASS","rationale":"After excluding already-disclosed product-specific calibration selection, the remaining claim is a falsifiable combination of explicit-key resolution, conflict abstention, inactive-map invariance, and re-entry reproducibility, each observable in stored predictions and routing records.","source_ids":["SRC3","SRC4"]},"bounded_next_test":{"status":"PASS","rationale":"A read-only replay of no more than 40 archived runs can compare the pooled baseline with isolated maps under correct, swapped, missing, and conflicting keys, then fit only formulation B and test locked formulation-A outputs and re-entry. This is bounded and consistent with calibration-domain validation practice.","source_ids":["SRC1","SRC2","SRC3"]},"no_obvious_safety_or_authority_stop":{"status":"PASS","rationale":"The proposed first test is offline, preserves approved calibrations and raw archives, and does not issue heater, endpoint, release, or DCS commands. Production authorization remains with process and quality owners. No obvious safety or authority stop applies provided incomplete provenance or irreproducible preprocessing causes the stated halt.","source_ids":["SRC3","SRC4"]}},"screen_survival":true,"world_novelty_boundary":"This bounded public-web screen found close generic process-spectroscopy routing and validation practices but no retained source disclosing the complete recipe-confirmation, inactive-map isolation, abstention, cross-map regression, and re-entry package for two filled-epoxy cure maps. That absence cannot establish world novelty, patentability, market size, expert acceptance, or realized value."}