{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"synchronized_release_dampening__computer_science","arm":"CONSTRAINED_HIGH","candidate_id":"synchronized_release_dampening__computer_science__CONSTRAINED_HIGH","decision":"PROPOSAL","abstention_reason":null,"proposal":{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp12_substrate_denial72_20260805","cell_id":"synchronized_release_dampening__computer_science","arm":"CONSTRAINED_HIGH","candidate_id":"synchronized_release_dampening__computer_science__CONSTRAINED_HIGH","proposal_index":1,"version":0,"title":"Passive Cam-Staged Restart for Computer-Rack Power Recovery","problem":"After a computer facility's AC supply returns, many rack power supplies can begin charging their input capacitors in the same short interval. Their overlapping inrush currents can exceed the transient service envelope of an otherwise adequately sized UPS, generator, feeder, or protective device, causing another interruption even when the eventual steady load is acceptable.","actors":["Rack power-supply banks","UPS, generator, or feeder bus","Branch contactors","Spring-return motorized cam sequencer","Facility electrical engineer","Computer services dependent on the racks"],"observable_state":"Following restoration of bus voltage, current-probe traces show branch-current rising edges clustered in one narrow interval, aggregate current crosses the source or protection envelope, and voltage collapses or protection opens; when measured after startup, the combined steady current remains within the continuous envelope.","consequence":"A restoration event can initiate repeated trip-and-restart cycles, extending the unavailability of the hosted computer systems and subjecting electrical components to repeated inrush events.","affected_objective":"Reliable recovery of computer capacity after power interruption while keeping transient feeder demand inside the electrical source's service envelope.","intervention":"Place rack branches behind normally open contactors operated by a spring-return, synchronous-motor cam sequencer. Loss of supply mechanically returns the cam to its home position. When supply returns, the motor rotates a shaft whose physically separated lobes close branch contactors in bounded cohorts; angular spacing is chosen to exceed the measured settling interval of each cohort's capacitor-charging inrush. A first lobe may energize one designated infrastructure cohort, while every remaining cohort receives a closure position within one revolution. Fuses, contactor ratings, mechanical interlocks, and a de-energizing stop protect the test installation. The sequencing action uses restored electrical energy, shaft motion, cam geometry, and contact separation rather than software or network commands.","structural_mapping":[{"archetype_element":"Shared Release Signal","domain_realization":"Return of usable AC voltage after an outage or transfer event."},{"archetype_element":"Waiting Population Boundary","domain_realization":"The set of de-energized rack power-supply banks connected to the recovering bus."},{"archetype_element":"Finite Choke Point","domain_realization":"The UPS, generator, feeder, transfer switch, or protective device with a bounded transient-current envelope."},{"archetype_element":"Release Correlation","domain_realization":"Temporal overlap among branch-current rising edges and capacitor-charging pulses."},{"archetype_element":"Dispersion Policy","domain_realization":"Physical angular offsets between cam lobes translate into separated branch-energization times."},{"archetype_element":"Admission Gate","domain_realization":"Normally open branch contactors prevent a cohort from drawing current until its cam position is reached."},{"archetype_element":"Fairness and Maximum-Wait Guard","domain_realization":"Every branch has a fixed cam position within one bounded revolution; the mechanism cannot repeatedly select one branch while omitting another."},{"archetype_element":"Priority Escape Lane","domain_realization":"One explicitly designated infrastructure cohort may occupy the first lobe without releasing the other cohorts simultaneously."}],"mechanism_mapping":[{"mechanism_slug":"cohort_based_reactivation","role":"Each cam lobe energizes a predefined, electrically bounded rack cohort at a distinct physical position in the rotation.","counterfactual_removal":"Replacing the separated lobes with one common closure position restores simultaneous branch energization and removes the proposed dispersion effect."},{"mechanism_slug":"semaphore_limited_release","role":"Cam geometry and contactor interlocking permit only the scheduled cohort's new inrush event during each angular interval, acting as a physical admission gate.","counterfactual_removal":"Hard-wiring the contactors to close directly on bus restoration bypasses the gate and allows all inrush events to overlap."}],"causal_chain":["AC restoration supplies the shared release event to all de-energized racks.","The spring-return cam begins from home and rotates at a mechanically fixed rate.","Separated lobes close cohort contactors at separated times rather than at the restoration instant.","Each cohort's capacitor-charging pulse is allowed to decay before the next scheduled rising edge.","Reduced temporal overlap lowers aggregate instantaneous current presented to the recovering electrical source.","If aggregate current remains inside the source and protection envelope, the electrical supply can remain available while all cohorts reach steady operation."],"baseline":"Use an otherwise identical comparator in which all branch contactors close directly when the test source is restored, preserving the same loads, source impedance, protection settings, and pre-test capacitor discharge state.","nearest_rivals":["Increase the transient-current rating of the UPS, generator, feeder, breaker, or transfer equipment without changing release timing.","Install a correctly sized passive inrush limiter or pre-charge network separately on every rack branch.","Use an electronic programmable power sequencer or server-firmware startup delay.","Redesign each power supply for lower capacitor-charging inrush."],"remaining_contrastive_claim":"The bounded claim for testing is that, when failure is caused by overlap among individually tolerable inrush pulses, physical separation of branch-closure times can reduce that overlap relative to simultaneous restoration without requiring greater source capacity or executable startup logic. It is not a claim about novelty, prevalence, demand, or a numerical effect size.","authority_safety":{"decision_authority":"A qualified facility electrical engineer controls any electrical test or installation; owners of hosted systems identify branches that must not be interrupted or delayed.","authorized_first_step":"Construct only an isolated, current-limited bench mock-up using low-voltage capacitor-input dummy loads, appropriately rated contactors, and guarded moving parts; do not connect the device to production racks or building mains.","excluded_actions":["Modification of energized building wiring","Connection to production UPS, generator, feeder, or rack circuits during the first evidence step","Bypassing overcurrent, grounding, interlock, or emergency-stop protection","Sequencing life-safety, fire-protection, medical, or other no-delay loads","Using contactors or cam hardware outside their voltage, current, interruption, or mechanical-duty ratings","Relying on a manual bypass that can close all test branches simultaneously"],"halt_rollback":"De-energize the isolated source immediately upon unexpected heating, arcing, contact welding, guard displacement, motor stall, repeated source limiting, or loss of branch order. Verify zero energy, discharge dummy-load capacitors through rated resistors, and remove the sequencer from the test circuit to restore the documented simultaneous-closure comparator."},"negative_tests":{"strongest_counterevidence":"A single rack cohort, energized alone from a fully recovered source, crosses the protection envelope or causes the same collapse; alternatively, the steady combined load exceeds the continuous envelope. Either result indicates inadequate capacity or per-branch inrush rather than primarily synchronized release.","problem_falsifier":"Matched measurements show no material temporal clustering of branch rising edges at restoration, or source failure occurs independently of whether branches are simultaneous or separated.","intervention_falsifier":"Instrumentation verifies that the cam produced non-overlapping branch-closure intervals, yet the matched test source still collapses or opens protection because of a single-cohort pulse, steady overload, harmonic interaction, motor demand, or another non-correlation mechanism.","risks":["A welded contactor could defeat staging and leave a branch energized unexpectedly.","A stalled or damaged cam could starve later cohorts and prolong computer-service unavailability.","Incorrect cohort sizing or insufficient angular separation could preserve harmful pulse overlap.","Rapid supply chatter could prevent full mechanical reset and create an unintended closure order.","Contact interruption can create arcing, heat, shock, and fire hazards if components or suppression are incorrectly selected.","Fixed ordering can delay storage, networking, cooling controls, or other dependencies needed before compute racks start.","The sequencer introduces a shared mechanical failure point into the restart path."]},"next_evidence_step":"On an isolated current-limited bench, connect four matched capacitor-input dummy loads through four contactors. Fully discharge the loads before each run. Perform ten simultaneous-closure runs and ten cam-sequenced runs with the same source impedance and protection setting. Use standalone current probes and an oscilloscope to record each branch rising edge, aggregate peak-current trace, bus-voltage trace, protection operation, cam reset, and completion of all four closures. The first decision is only whether verified temporal separation changes source collapse or protection operation under matched conditions.","prior_art_status":"UNSEARCHED","diversity_from_prior_proposals":"Not assessed against other proposals because runtime isolation forbids inspecting them. Within the supplied packet, this candidate realizes the archetype as an electromechanical computer-power recovery mechanism rather than as software retry, cache, queue, or reporting logic.","revision_record":{"parent_version":null,"progress_targets_addressed":["Construct an independently recognizable computer-systems problem","Preserve synchronized-release causal structure","Make the essential intervention substrate-compliant","Specify serious rivals, safeguards, falsifiers, and a bounded evidence step"],"conceptual_changes":["Initial candidate formulation maps synchronized waiters to de-energized rack power supplies and the finite choke to transient electrical-source capacity.","Release dispersion is implemented through cam geometry and contactor motion rather than computation."],"operational_changes":["Defined an isolated four-branch bench comparison with controlled capacitor discharge, source impedance, and protection settings.","Excluded production, energized, life-safety, and protection-bypass work from the first step."],"evidence_changes":["Prior art remains unsearched under the closed-book constraint.","Specified direct electrical and timing measurements capable of separating correlation failure from single-load or steady-capacity failure."],"claim_changes":["Restricted the contrastive claim to temporally overlapping, individually tolerable inrush pulses.","Made no claim of novelty, prevalence, demand, or numerical effect size."]},"substrate_contract":{"primary_allowed_process":"PHYSICAL_MATERIAL","counterfactual_independence":"After installation and arming, removal of software, algorithmic inference, databases, dashboards, reporting, incentives, authorization logic, and procedural enforcement does not remove the essential effect. Restored electrical energy turns a motor; shaft rotation and spatially offset solid cam lobes actuate normally open contactors at different physical times. The resulting separation of current pulses persists without a processor, network, sensor-analysis path, operator decision, or software control loop.","forbidden_channel_audit":"No firmware command, application signal, model, scheduler, telemetry pipeline, or human release procedure performs branch admission. Measurement instruments are used only to test the causal claim and are not in the operative path. Engineering approval, lockout, guarding, and inspection are safety wrappers; they neither generate the staggered timing nor substitute for the cam-and-contactor mechanism."}}}