{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"preimage_set_characterization__computer_science","arm":"CONSTRAINED_HIGH","candidate_id":"preimage_set_characterization__computer_science__CONSTRAINED_HIGH","decision":"PROPOSAL","abstention_reason":null,"proposal":{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"preimage_set_characterization__computer_science","arm":"CONSTRAINED_HIGH","candidate_id":"preimage_set_characterization__computer_science__CONSTRAINED_HIGH","proposal_index":1,"version":0,"title":"Galvanically Isolated Branch-Trip Preimage Cradle","problem":"In a modular compute chassis, a common power-supply protection trip is a many-to-one output: a hard short in any one feeder branch, several simultaneous branch faults, or an upstream supply or harness defect can produce the same shutdown. Repeatedly disconnecting modules may identify an example but can obscure multiplicity and does not establish that every declared branch was tested under the same stimulus.","actors":["hardware repair technician","laboratory electrical-safety authority","modular compute chassis","power-supply and distribution harness","compute modules and their feeder branches","diagnostic cradle"],"observable_state":"When the assembled chassis is energized, the shared supply immediately enters overcurrent protection. The chassis output alone does not reveal which module feeder branches are individually capable of reproducing a specified trip-like current condition.","consequence":"The chassis remains unavailable, repeated full-power trials can stress connectors or components, and a single module may be replaced as the presumed cause while another qualifying branch remains undiscovered.","affected_objective":"Bound the complete set of declared module feeder branches that individually satisfy a reproducible hard-short predicate before component-level diagnosis or replacement.","intervention":"Insert an energy-limited diagnostic cradle between the de-energized chassis harness and its modules. The cradle fans every declared feeder into a separate galvanically isolated low-voltage DC test channel. Each channel applies the same bounded stimulus and contains a calibrated magnetic current-threshold element that mechanically latches a numbered flag when that branch sustains current above the predicate threshold for the specified dwell interval. Every branch is stimulated independently, so one short cannot starve or mask another. The simultaneously visible flag array is the physical preimage set: all numbered branches in the declared connector domain that map to the target predicate. Open-circuit and short-circuit reference positions provide physical checks of channel coverage and threshold behavior. The result identifies qualifying feeder branches, not the failed component or the historical cause of the original shutdown.","structural_mapping":[{"archetype_element":"Mapping Under Review","domain_realization":"For each declared feeder branch b, the fixed mapping T(b) is whether that branch, under the cradle's stated voltage, current limit, polarity, and dwell time, actuates its mechanical threshold flag."},{"archetype_element":"Output Condition or Target Value","domain_realization":"The target output is T(b)=true: sustained branch current crosses the calibrated hard-short threshold and physically latches the flag."},{"archetype_element":"Input Domain Boundary","domain_realization":"The domain is the numbered set of module feeder contacts present on the selected chassis harness; upstream PSU internals, signal pins, chassis-frame leakage, and multi-branch interaction states are outside it."},{"archetype_element":"Preimage Membership Rule","domain_realization":"A branch is included if and only if its own isolated channel latches during the bounded stimulus and the flag remains latched after stimulus removal."},{"archetype_element":"Candidate Input Enumeration","domain_realization":"A one-channel-per-feeder physical fan-out tests every member of the declared domain and displays membership at the corresponding numbered position."},{"archetype_element":"Collision and Multiplicity Check","domain_realization":"Independent supplies and simultaneously retainable flags allow zero, one, or several qualifying branches to remain visible without collapsing them into the common chassis-trip output."},{"archetype_element":"Completeness Evidence","domain_realization":"Connector-contact continuity, a numbered channel count matched to the harness drawing, and open/short reference positions test whether every declared feeder is connected and whether both exclusion and inclusion responses are physically produced."},{"archetype_element":"Boundary Case Register","domain_realization":"Near-threshold resistance, capacitive inrush, intermittent contact, polarity dependence, temperature dependence, semiconductor turn-on, and faults that appear only at operating voltage remain explicitly outside or uncertain under the fixed test condition."},{"archetype_element":"Downstream Use Guardrail","domain_realization":"Flag membership licenses branch isolation and further bench examination only; it does not establish the failed component, exclude upstream defects, or prove which branch caused the historical event."}],"mechanism_mapping":[{"mechanism_slug":"predicate_satisfaction_filter","role":"Each calibrated magnetic threshold element converts the physical current predicate into a retained mechanical included-or-excluded state without computation.","counterfactual_removal":"Without the threshold element, current could flow but the fixture would not physically apply the fixed membership predicate."},{"mechanism_slug":"preimage_table","role":"The spatially indexed array of numbered latching flags represents the entire discovered preimage as a simultaneously inspectable physical set.","counterfactual_removal":"Without the array, separate channel responses would not remain associated with all domain members, and multiplicity could again collapse into a single trip observation."},{"mechanism_slug":"collision_analysis_matrix","role":"Galvanic separation prevents one branch's voltage collapse from masking other branches and preserves simultaneous multiple membership.","counterfactual_removal":"With a shared diagnostic rail, one hard short could collapse the stimulus and hide additional qualifying branches."},{"mechanism_slug":"coverage_completeness_audit","role":"One physical channel per connector contact plus open and short reference positions checks domain coverage and endpoint response.","counterfactual_removal":"Without contact coverage and reference checks, an unconnected or insensitive channel could be mistaken for a valid exclusion, defeating the bounded completeness claim."}],"causal_chain":["The original shared supply maps several possible branch-fault states to one indistinguishable protection-trip output.","The cradle physically separates the declared feeder domain into independently energized channels.","The same bounded electrical stimulus is applied to every feeder without allowing one feeder to suppress another's test voltage.","A qualifying low-impedance branch carries enough sustained current to actuate its channel's magnetic threshold element.","The element latches a durable mechanical flag at the spatial position assigned to that branch.","All latched positions remain visible together, preserving multiplicity and forming the bounded preimage of the target predicate.","Reference positions and contact-continuity checks bound the completeness claim to functioning channels and the declared connector domain.","Further diagnosis proceeds on every included branch while upstream, interaction-only, and voltage-dependent rivals remain open."],"baseline":"Disconnect one module at a time and retry the chassis on its normal shared supply. This can disturb an intermittent fault, subjects the system to repeated operating-power events, allows the first discovered branch to stop the search, and provides no retained simultaneous representation of multiple qualifying branches.","nearest_rivals":["Test each module sequentially with a conventional current-limited bench supply; this applies a comparable physical predicate but depends on repeated reconnection and does not retain a simultaneous full-domain set unless results are separately recorded.","Inject current into the shared rail and inspect heating with a thermal camera; this can localize dissipation but may miss masked branches, confuse normal thermal mass with membership, and does not intrinsically certify coverage of every feeder.","Install electronic current-sense channels with telemetry and analytics; this can measure branch currents but makes downstream electronics, data handling, or software the operative reverse-mapping layer.","Use an ohmmeter at each disconnected feeder; this is simple but serial, contact-sensitive, and may not reproduce the sustained-current threshold that defines the target predicate.","Use a combinatorial load fixture to test branch subsets; this addresses interaction-only trips but has a different, much larger input domain than the proposed singleton-feeder preimage."],"remaining_contrastive_claim":"For the explicitly bounded singleton-feeder domain, the cradle combines identical isolated physical stimulation, retained per-branch threshold flags, and one-to-one connector coverage so that all branches satisfying the stated hard-short predicate remain visible together without software. It does not characterize combinations of individually subthreshold branches or claim that a qualifying branch caused the historical trip.","authority_safety":{"decision_authority":"The laboratory electrical-safety authority approves stimulus limits and connector adaptation; the hardware repair lead decides whether a flagged module proceeds to component-level examination.","authorized_first_step":"On a de-energized non-production chassis or representative harness, verify connector identity and polarity, set voltage and energy limits below module damage thresholds, populate open and short reference positions, and run one bounded cradle trial behind existing protective isolation.","excluded_actions":["connecting the cradle to mains conductors","using normal chassis operating power as the diagnostic stimulus","bypassing fuses, isolation, current limits, interlocks, or protective earth","connecting storage media or modules whose safe diagnostic-voltage envelope is unknown","treating a latched flag as proof of a particular failed component or historical cause","returning a chassis to service solely from this test"],"halt_rollback":"Disconnect the isolated source immediately upon unexpected heating, odor, arcing, connector discoloration, reference-position failure, or stimulus instability. Allow thermal components to cool, reset mechanical flags, remove the cradle, restore the original de-energized harness configuration, and quarantine any visibly affected connector before further testing."},"negative_tests":{"strongest_counterevidence":"The original protection trip follows the PSU or upstream harness when all module feeders are disconnected, or occurs only when several individually normal branches operate together; either observation undercuts a singleton-branch explanation.","problem_falsifier":"Independent conventional measurements show that every declared module branch remains outside the hard-short predicate while an upstream supply defect, harness short, or aggregate inrush fully accounts for the common trip.","intervention_falsifier":"The cradle fails to latch for a reference short, latches for an open or known-good reference, omits a declared feeder contact, masks one of two simultaneous reference shorts, or gives materially different membership under repeated unchanged stimuli.","risks":["A miswired adapter or reversed polarity could damage a module.","Even limited test energy can heat a shorted trace, connector, or magnetic element.","Capacitive charging or normal low-resistance loads could create false inclusion.","A semiconductor fault requiring operating voltage, temperature, or control signals could be falsely excluded.","Intermittent contact may change when the harness is moved into the cradle.","Galvanic isolation or channel calibration drift could create unequal predicates.","Users may mistake predicate membership for component-level causation or completeness beyond the declared domain."]},"next_evidence_step":"Build a non-production two-reference, four-branch bench coupon representing open, known-good load, one hard short, and two simultaneous hard shorts. Under a predeclared voltage, current limit, polarity, dwell time, and temperature band, verify physical inclusion, exclusion, simultaneous retention, channel isolation, contact coverage, and safe energy dissipation. Stop before testing production modules if any reference response is incorrect or if the bounded stimulus exceeds the coupon's thermal limit.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Not assessed against other proposals because runtime isolation forbids inspecting them; this candidate is internally distinguished by a galvanically isolated, electromechanical, software-independent measurement substrate in modular computing hardware.","revision_record":{"parent_version":null,"progress_targets_addressed":["Preserve the archetype's many-to-one reverse-membership structure.","Make the essential intervention a physical measurement process.","Bound completeness to an explicit input domain.","Separate predicate membership from causal attribution.","Specify authority, rollback, falsifiers, and a limited first experiment."],"conceptual_changes":["Initial version defines the preimage over singleton feeder branches rather than unbounded combinations of chassis states.","Initial version treats the common trip as the motivating lossy output and a fixed hard-short predicate as the reproducible mapping under review."],"operational_changes":["Initial version uses isolated low-energy channels and mechanically retained indicators.","Initial version includes physical open/short references and one-to-one connector coverage."],"evidence_changes":["Prior art remains unsearched under the closed-book constraint.","All expected performance remains a testable hypothesis pending the bounded coupon trial."],"claim_changes":["No novelty, prevalence, demand, or effect-size claim is made.","Completeness is restricted to declared feeders, functioning channels, and the fixed diagnostic stimulus."]},"substrate_contract":{"primary_allowed_process":"MEASUREMENT_INSTRUMENTATION","counterfactual_independence":"Removing software, algorithmic inference, databases, dashboards, telemetry, reporting, incentives, authorization rules, and procedural enforcement leaves the essential effect intact: isolated electrical energy still flows through every feeder, qualifying current still actuates its calibrated magnetic element, and numbered mechanical flags still retain the complete observed membership pattern. Human setup and safeguards constrain use but do not create the reverse mapping.","forbidden_channel_audit":"The proposed cradle contains no microcontroller, programmable logic, software control loop, model, database, network connection, telemetry path, recommender, or digital report. Its membership decision is produced by current, magnetic force, spring latching, galvanic separation, and spatial correspondence. Governance determines who may conduct the test but is not the intervention. If the branch set required analytics or human recordkeeping to exist, the proposal would fail this contract; the retained physical flag array prevents that dependency."}}}