{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"context_keyed_representation_switching__accounting_auditing","arm":"CONSTRAINED_HIGH","candidate_id":"context_keyed_representation_switching__accounting_auditing__CONSTRAINED_HIGH","decision":"PROPOSAL","abstention_reason":null,"proposal":{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"context_keyed_representation_switching__accounting_auditing","arm":"CONSTRAINED_HIGH","candidate_id":"context_keyed_representation_switching__accounting_auditing__CONSTRAINED_HIGH","proposal_index":1,"version":0,"title":"Mechanically Tank-Keyed Cam Gauge for Multi-Tank Inventory Counts","problem":"During a physical inventory count, one portable level gauge may be used across several irregularly shaped liquid-storage tanks. Each tank has a distinct level-to-volume calibration curve, so the same measured liquid height can represent different inventory quantities. A generic ruler followed by selection of a tank-specific table permits the correct level measurement to be converted through the wrong tank map, producing an incorrect quantity even when the level reading and arithmetic are individually correct.","actors":["Physical inventory counter","Independent auditor","Tank custodian","Portable mechanical level gauge","Storage tanks and their liquid contents"],"observable_state":"The directly observable state is the liquid level transmitted through the gauge tape or float, the physical key geometry of the tank-port docking collar, the mechanically exposed tank identifier, and the resulting analog volume indication.","consequence":"Engaging a calibration curve belonging to another tank converts a valid level reading into the wrong physical inventory quantity, which can propagate into inventory reconciliation or valuation records.","affected_objective":"Obtain a traceable, tank-appropriate physical quantity measurement for inventory existence and quantity testing without silently applying another tank's calibration representation.","intervention":"Equip one portable level-measurement head with a protected magazine of persistent, machined calibration cams, one cam per tank. Each tank port receives a unique passive docking collar whose mechanical key moves the gauge selector into engagement with only that tank's cam. Full seating engages an over-center detent, exposes a mechanically linked tank-ID flag, and opens the quantity dial; a missing, partial, or unknown key leaves the dial shuttered. Level movement drives the selected cam, whose profile physically transforms level displacement into a volume indication. All other cams remain disengaged and protected, so redocking at a previously measured tank reinstates its unchanged map.","structural_mapping":[{"archetype_element":"Shared Substrate Scope","domain_realization":"One portable tape-or-float measurement head, drive train, follower, pointer, and enclosure are shared across the bounded set of tanks."},{"archetype_element":"Representation Portfolio","domain_realization":"A magazine of individually identified machined cams stores the distinct level-to-volume curve for each tank."},{"archetype_element":"Context Key Space","domain_realization":"The allowable contexts are the enumerated tank ports, each represented by a mutually incompatible docking-collar geometry; unmatched geometry is the unknown context."},{"archetype_element":"Context Cue and Selection Rule","domain_realization":"Physical collar geometry displaces a selector fork to the corresponding cam without inferred identity or operator table choice."},{"archetype_element":"Active Representation Indicator","domain_realization":"A flag mechanically linked to the selector exposes the engaged tank identifier only after full docking."},{"archetype_element":"Representation Isolation Boundary","domain_realization":"Only one cam contacts the follower; inactive cams are mechanically disengaged, covered, and unaffected by ordinary measurement motion."},{"archetype_element":"Switch Transition Guard","domain_realization":"A full-seat interlock and over-center detent prevent dial exposure during partial engagement or between-cam positions."},{"archetype_element":"Unknown-Context Fallback","domain_realization":"An unrecognized or incompletely seated collar cannot open the output shutter, so the device produces no plausible volume reading."},{"archetype_element":"Re-Entry Integrity Test","domain_realization":"Reference-height stops permit the indication from each cam to be checked before and after repeated excursions through other tank contexts."},{"archetype_element":"Shared Invariant Layer","domain_realization":"The level-sensing tape, drive ratio, follower datum, pointer, and units remain common; only the tank-specific nonlinear calibration profile switches."}],"mechanism_mapping":[{"mechanism_slug":"context_to_map_routing_table","role":"The set of collar keyways and selector-fork positions physically encodes which tank engages which calibration cam.","counterfactual_removal":"Without the keyed routing geometry, the operator could engage an arbitrary cam and the context-specific selection effect would be lost."},{"mechanism_slug":"per_context_model_checkpoint","role":"Each fixed cam profile is a persistent physical checkpoint of one tank's level-to-volume mapping.","counterfactual_removal":"Without separately preserved cams, switching tanks would require overwriting, rebuilding, or externally looking up the prior mapping."},{"mechanism_slug":"active_map_status_indicator","role":"A selector-linked flag displays the identity of the cam actually in contact with the follower.","counterfactual_removal":"Conversion would still occur, but the engaged representation could not be directly checked against the docked tank."},{"mechanism_slug":"hysteresis_and_debounce_filter","role":"The over-center detent requires complete docking before engagement and resists vibration-driven movement between adjacent cam positions.","counterfactual_removal":"Partial seating or vibration could create mixed contact or unstable switching."},{"mechanism_slug":"safe_default_map_fallback","role":"A spring-return shutter blocks the dial unless a recognized key reaches a valid detent.","counterfactual_removal":"An unknown or partial context could expose a plausible output from a stale or intermediate cam."},{"mechanism_slug":"context_reinstatement_protocol","role":"Undocking disengages the current cam; redocking through a previously used collar mechanically re-engages the preserved cam at its fixed datum.","counterfactual_removal":"Return to an earlier tank would not reliably restore its mapping without manual reconstruction."},{"mechanism_slug":"map_difference_and_integrity_check","role":"Gauge-pin reference positions and engraved cam identifiers allow each stored profile's datum and selected points to be checked physically.","counterfactual_removal":"Cam wear, displacement, or substitution would be harder to distinguish from ordinary measurement variation."}],"causal_chain":["A tank's passive docking collar enters the portable gauge socket.","The collar's unique geometry shifts the selector toward the corresponding stored calibration cam.","Only complete seating closes the detent, engages that cam, exposes its tank-ID flag, and opens the dial shutter.","The tape or float converts liquid level into mechanical drive displacement.","The selected cam profile converts that displacement through its tank-specific nonlinear level-to-volume correspondence.","The follower and pointer expose a direct volume reading while all inactive cams remain mechanically disconnected.","Undocking and later redocking at another or prior tank selects the new or reinstated persistent map without altering the others.","The context-appropriate physical quantity can then be used as evidence in the inventory count and reconciliation."],"baseline":"Use a generic dip tape or ruler to measure level, then have the counter select a tank-specific paper calibration table and perform or transcribe the conversion. Serious alternative baselines include a separate permanently calibrated gauge for every tank and a portable gauge with a manually selectable scale plate.","nearest_rivals":["A dedicated, fixed calibrated level gauge on every tank, which avoids map switching by duplicating the full measurement instrument.","A generic dip measurement followed by a tank-specific paper strapping table, which preserves distinct maps but relies on external selection and transcription.","A portable gauge with a manually indexed cam or interchangeable scale plate, which shares the instrument but does not physically bind map selection to the tank port.","Direct weighing of the vessel or transferred liquid, which measures inventory through a different physical quantity and may avoid level-to-volume maps."],"remaining_contrastive_claim":"Relative to paper tables or manually selected scales, the proposed device physically binds the active level-to-volume representation to the tank-port context and blocks output for an unmatched context; relative to dedicated per-tank gauges, it preserves distinct persistent mappings while sharing the level-sensing and display substrate.","authority_safety":{"decision_authority":"The tank owner controls vessel access and hardware installation; the audit engagement lead determines whether prototype readings may be considered as audit evidence; the site safety authority controls work near hazardous or pressurized tanks.","authorized_first_step":"Authorize only a detached bench prototype using surrogate collars, simulated level travel, and non-live cam profiles; do not connect it to an operating vessel or financial record.","excluded_actions":["Opening, modifying, or entering a live or pressurized tank","Using the prototype in a hazardous atmosphere without separate equipment approval","Posting ledger adjustments from prototype readings","Replacing required certified custody-transfer or safety instruments","Changing tank calibration data without an independently controlled calibration basis"],"halt_rollback":"Stop testing if any recognized collar engages the wrong cam, any unknown collar exposes a reading, the selector can rest between cams, or a cam shifts outside the predeclared dimensional tolerance. Remove the prototype and continue using the existing measurement method; no accounting record is changed during the first step."},"negative_tests":{"strongest_counterevidence":"The proposal has little value if each tank already has a reliable, physically noninterchangeable dedicated gauge, or if the candidate tanks' level-to-volume curves are equivalent within the accepted measurement tolerance and therefore do not require persistent distinct maps.","problem_falsifier":"Review of blinded historical or simulated conversions shows no wrong-tank-map exposure, or dimensional characterization shows that one common mapping is adequate across the bounded tank set.","intervention_falsifier":"In randomized bench trials, any fully seated collar selects the wrong cam, any unknown or partial collar yields a plausible volume, repeated switching changes an inactive cam's output, or re-entry error exceeds the predeclared mechanical tolerance.","risks":["A collar could be installed on the wrong tank, making a mechanically consistent but semantically wrong pairing.","Wear, debris, corrosion, or deformation could alter key engagement or cam-follower motion.","Tank modification, settlement, or recalibration could make a preserved cam obsolete while its identifier remains valid.","The shared measurement head creates a common-mode failure across all included tanks.","The visible interlock could create false confidence if liquid level sensing, temperature effects, or the underlying tank calibration is wrong.","Using one portable head across tanks could transfer contamination unless materials and cleaning controls are separately established.","Fine cam geometry may be difficult to manufacture or inspect to the required measurement tolerance."]},"next_evidence_step":"Build a bench-only three-context prototype with deliberately distinct cam profiles, three matching collars, one unknown collar, and a simulated level actuator. Before testing, declare allowable selection and dimensional errors. Randomize repeated sequences across five reference levels per map, include partial insertion and unknown-key challenges, then compare indicated volumes with the cam design values and repeat the first map after intervening switches to test isolation and re-entry. The step ends with recorded mechanical observations and no live-tank or ledger use.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Not assessed against other proposals because runtime isolation prohibits inspecting them; this candidate was derived only from the supplied archetype and domain card.","revision_record":{"parent_version":null,"progress_targets_addressed":["Initial construction under the binding constrained-substrate requirement","Concrete accounting-and-auditing measurement problem","Explicit counterfactual independence from forbidden wrappers"],"conceptual_changes":["None; initial version."],"operational_changes":["None; initial version."],"evidence_changes":["Prior art remains unsearched; only a bounded bench evidence step is proposed."],"claim_changes":["No novelty, prevalence, demand, or effect-size claim is made."]},"substrate_contract":{"primary_allowed_process":"MEASUREMENT_INSTRUMENTATION","counterfactual_independence":"If all software, algorithmic inference, reporting, incentives, authorization rules, and procedural enforcement are removed, a correctly installed physical collar still selects one cam, the measured level still drives that cam, the cam still transforms displacement into the corresponding volume indication, inactive cams remain disengaged, and an unmatched collar still leaves the shutter closed. Those wrappers can document or govern use but are not required for the essential context selection and measurement transformation.","forbidden_channel_audit":"The intervention contains no computation, database, dashboard, recommender, learned selector, information-routing system, or software control loop. It does not depend on training, incentives, review, permission, or a human lookup procedure to choose the representation. The context cue, map bank, selector, isolation, transition guard, fallback, conversion, and indicator are passive or mechanically driven instrument components. Human authority is limited to safe testing and evidentiary use, not the causal measurement operation."}}}