{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"search_space_pruning__computer_science","arm":"CONSTRAINED_HIGH","candidate_id":"search_space_pruning__computer_science__CONSTRAINED_HIGH","decision":"PROPOSAL","abstention_reason":null,"proposal":{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"search_space_pruning__computer_science","arm":"CONSTRAINED_HIGH","candidate_id":"search_space_pruning__computer_science__CONSTRAINED_HIGH","proposal_index":1,"version":0,"title":"Calibrated Phase-Change Mosaic for Pruning PCB Fault-Location Regions","problem":"When a densely routed computing-system circuit board has an intermittent resistive short or leakage fault, the possible defect sites span many traces, vias, packages, and board regions. Point-by-point electrical or thermal examination of every site can be costly and may disturb the fault before it is localized.","actors":["Hardware failure-analysis engineer","Faulty printed circuit board","Resistive defect site","Current-limited excitation source","Replaceable phase-change indicator mosaic","Calibration reference elements"],"observable_state":"After a bounded diagnostic current pulse, paired low- and high-threshold indicator patches above each board region exhibit persistent, directly visible phase states. Transitioned calibration references establish that the intended thermal exposure occurred; transitioned, borderline, and unchanged regional patches form a spatial heat-signature map without computational interpretation.","consequence":"Without a defensible early screen, the investigator must examine a large set of possible sites or may narrow informally and overlook the actual fault region, increasing diagnostic effort and the opportunity for invasive probing damage.","affected_objective":"Reduce the number of PCB regions requiring detailed fault-location work while keeping the true region reachable and limiting thermal and false-exclusion risk.","intervention":"Place a replaceable, electrically insulating mosaic of paired phase-change threshold patches over the de-energized board, together with reference patches coupled to calibration resistors. Apply one current-limited, energy-bounded pulse through the suspect power domain. A region may be physically marked as pruned only when its low-threshold patch remains unchanged while the adjacent reference confirms adequate exposure and the specified fault class would necessarily have produced enough local heat to cross that threshold. Retain all transitioned and ambiguous regions, boundary neighbors, and a prespecified sample of apparently cold regions. The persistent mosaic itself records the screen. A fresh mosaic, a different safe pulse condition, or direct probing provides reentry.","structural_mapping":[{"archetype_element":"Search Space Definition","domain_realization":"All accessible board regions intersecting the suspect electrical domain constitute the initial set of possible resistive-fault locations."},{"archetype_element":"Candidate or Region Representation","domain_realization":"Each labeled mosaic tile represents a bounded physical board region; boundary tiles overlap neighboring regions to avoid gaps."},{"archetype_element":"Objective or Success Criterion","domain_realization":"Retain every region physically compatible with the specified heat-producing fault while reducing regions sent to invasive probing."},{"archetype_element":"Constraint Filter","domain_realization":"A cold region is excludable only if calibration confirms sufficient exposure and the hypothesized fault's minimum heat release exceeds the tile's characterized transition threshold."},{"archetype_element":"Feasibility Test","domain_realization":"Paired threshold patches test whether local heat was absent, present, or too close to the measurement boundary for exclusion."},{"archetype_element":"Pruning Rule","domain_realization":"Prune only calibrated, untransitioned regions; never prune a transitioned, borderline, uncovered, poorly coupled, or calibration-invalid region."},{"archetype_element":"Diversity Preservation Rule","domain_realization":"Keep all boundary neighbors plus representative cold tiles from different board constructions and component-density zones for direct checking."},{"archetype_element":"False-Negative Review","domain_realization":"Probe the retained cold-tile sample and compare mosaic classifications with seeded-fault ground truth during validation."},{"archetype_element":"Reentry or Exception Path","domain_realization":"Reopen a region using a fresh lower-threshold mosaic, changed pulse polarity within the safe envelope, or direct electrical probing when the fault model or calibration changes."},{"archetype_element":"Pruning Audit Trail","domain_realization":"Irreversible color or phase states, printed tile identifiers, reference states, and the retained used sheet preserve the physical basis for each exclusion."},{"archetype_element":"Confidence Threshold","domain_realization":"The paired indicators create a physical bracket: ambiguous partial transitions cause retention rather than exclusion."},{"archetype_element":"Criteria Update Trigger","domain_realization":"Reference failure, unexpected heat diffusion, a changed fault hypothesis, or disagreement with a sampled direct probe invalidates affected exclusions."}],"mechanism_mapping":[{"mechanism_slug":"constraint_filtering","role":"Physically screens out only regions incompatible with the calibrated minimum heat signature of the specified resistive-fault class.","counterfactual_removal":"Without this calibrated cold-region exclusion, the mosaic may display heat but does not prune the location search."},{"mechanism_slug":"triage_filter","role":"Separates regions into retained-hot, retained-ambiguous, audit-holdout, and physically justified cold exclusions before invasive examination.","counterfactual_removal":"Removing this separation sends every region to the same costly downstream examination and eliminates tractability."},{"mechanism_slug":"sample_audit_of_exclusions","role":"Preserves representative cold regions for direct probing to expose threshold, coupling, or fault-model false negatives.","counterfactual_removal":"The physical screen still narrows the space, but there is no bounded check against systematic over-pruning."}],"causal_chain":["A resistive short or leakage site converts part of the bounded electrical pulse into localized heat.","Thermal energy conducted toward the board surface reaches the overlying threshold patches.","Patches whose transition energy is exceeded undergo a persistent visible material phase change.","Paired thresholds and calibration references distinguish adequate cold evidence from inadequate exposure or borderline heating.","Only calibrated cold regions are removed from immediate detailed investigation; hot, ambiguous, neighboring, and sampled cold regions remain reachable.","Detailed electrical or microscopic examination is concentrated on the retained physical regions while the used mosaic preserves the exclusion basis."],"baseline":"Examine candidate sites sequentially with point probes, microscopy, or repeated whole-board thermal observations, with region selection based on investigator judgment and no persistent simultaneous threshold record.","nearest_rivals":["Infrared-camera imaging of the board during excitation, which offers richer continuous imagery but depends on an active optical instrument and may require interpretation of transient frames.","Time-domain reflectometry or impedance localization, which measures electrical discontinuity by distance but may be difficult to map onto branched, multilayer geometry.","Manual divide-and-test isolation of nets or board sections, which can localize faults but requires repeated disconnection or probing.","Dense bed-of-nails electrical testing, which provides direct measurements but requires accessible test points and a board-specific fixture."],"remaining_contrastive_claim":"For fault classes with a characterized minimum local heat release and a validated thermal path to the surface, a passive paired-threshold mosaic can supply simultaneous, persistent physical evidence for excluding calibrated cold board regions before detailed probing; unlike sequential electrical isolation, its essential discrimination is produced by material phase transitions across many regions during one bounded excitation. This is a testable contrast, not a novelty or performance claim.","authority_safety":{"decision_authority":"A qualified hardware failure-analysis engineer with responsibility for the board and authority to set its diagnostic energy envelope.","authorized_first_step":"Run a coupon-only bench trial using a current-limited source, nonfunctional board materials, calibration resistors, and seeded resistive faults; do not begin on an operational asset.","excluded_actions":["Exciting an energized production system","Testing battery, pyrotechnic, medical, life-safety, or unknown-energy circuits","Exceeding component, insulation, or board thermal limits","Treating an uncalibrated or partially transitioned cold tile as exclusion evidence","Discarding the used mosaic before discrepancies are resolved"],"halt_rollback":"Stop excitation immediately upon current-limit instability, unexpected heating, odor, smoke, delamination, reference failure, or indicator contamination. Disconnect the source, allow the coupon to cool, quarantine it for inspection, and invalidate all exclusions from that run. Because the indicator transition is irreversible, rollback consists of discarding the sheet and repeating only with a fresh mosaic after the cause is resolved."},"negative_tests":{"strongest_counterevidence":"The actual fault may dissipate too little heat, conduct heat to a remote structure, occur only outside the imposed pulse condition, or be obscured by spreading from a healthy hot component; any of these would break the assumed mapping between defect location and tile transition.","problem_falsifier":"The candidate problem is falsified for this intervention if the suspect set is already small enough for cheap direct examination, or if the relevant failure class has no reproducible local thermal signature under a safe excitation envelope.","intervention_falsifier":"In blinded seeded-fault coupon trials, the intervention is falsified if any region containing an in-scope fault is classified as pruned under valid calibration, or if calibration and paired thresholds cannot reproducibly separate insufficient exposure from defensible cold evidence.","risks":["Thermal or electrical damage from diagnostic excitation","False exclusion caused by poor surface coupling, threshold variation, or an incorrect minimum-heat model","Mislocalization caused by lateral heat spreading or remote heat sinking","Electrostatic discharge or contamination introduced by the overlay","Loss of spatial coverage over inaccessible packages or board surfaces","Operator exposure to hot surfaces or phase-change material if containment fails"]},"next_evidence_step":"Fabricate one gridded nonfunctional PCB coupon with calibration resistors and twelve blinded seeded conditions spanning in-scope resistive faults, no-fault controls, boundary faults, and thermally shunted faults. Prespecify the pulse-energy ceiling, tile thresholds, calibration-validity rule, pruning rule, and direct-probe ground truth. Accept further investigation only if no seeded in-scope fault region is pruned and every exclusion is reconstructible from the retained sheet; otherwise revise or abandon the physical mapping.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Not assessed against other proposals because runtime isolation forbids inspecting them; within this candidate, the distinctive substrate is a passive, spatially parallel calorimetric phase-change measurement rather than software search, analytics, or governance.","revision_record":{"parent_version":null,"progress_targets_addressed":["Construct one causally defensible candidate under the binding substrate constraint","Preserve explicit search-space pruning structure","Specify false-negative safeguards and reentry","Provide bounded physical validation and falsifiers"],"conceptual_changes":["Translated candidate-space pruning into spatial exclusion of PCB fault-location regions","Made absence of a calibrated thermal signature the hard physical pruning criterion"],"operational_changes":["Added paired thresholds, calibration references, overlapping boundary tiles, and retained cold-region samples","Restricted first use to a seeded nonfunctional coupon"],"evidence_changes":["Set prior-art status to unsearched","Defined a blinded ground-truth coupon trial as the first evidence step"],"claim_changes":["Limited the claim to fault classes with characterized heat release and validated surface coupling","Made no claim of novelty, prevalence, demand, or effect size"]},"substrate_contract":{"primary_allowed_process":"MEASUREMENT_INSTRUMENTATION","counterfactual_independence":"The essential effect is the conversion of spatially distributed heat into persistent material phase states under a calibrated excitation. If all software, algorithms, databases, dashboards, reports, incentives, authorization rules, and procedural enforcement are removed, the mosaic still undergoes the same local transitions and physically distinguishes heated, cold, and ambiguous regions. Human or computational wrappers may label or archive the result, but they do not create the discriminating signal.","forbidden_channel_audit":"The proposal contains no model, inference algorithm, software control loop, database, recommender, or information-routing system. The pulse source requires only a fixed physical current and energy limit; it need not be software-controlled. Safety authority and handling instructions constrain use but are not the mechanism that localizes the fault. The persistent indicator sheet—not downstream analytics, reporting, incentives, or workflow—provides the operative pruning evidence."}}}