{"schema_version":1,"experiment_id":"eoa_inverse_innovation_exp03_full320_20260801","cell_id":"negative_space_design__operations_research","trajectory_id":"R","attempt_index":0,"archetype_slug":"negative_space_design","domain_slug":"operations_research","decision":"CANDIDATE","problem_id":"brittle_high_utilization_rolling_horizon_schedules","causal_lever_id":"protected_recovery_capacity_with_release_triggers","proposal":{"problem":"Rolling-horizon schedules optimized near nominal full utilization can become brittle under variable arrivals, processing times, breakdowns, or rework: small disruptions propagate across jobs, increasing lateness, overtime, and costly rescheduling even when average demand appears feasible.","actors_substrate":["planning analyst or scheduling team","dispatchers and frontline operators","customers or downstream processes awaiting jobs","queued jobs with priorities and due dates","finite machines, crews, vehicles, or service capacity","rolling-horizon optimization model and operational telemetry"],"observable_state":"Schedules repeatedly assign nearly every capacity interval; minor deviations consume the few incidental gaps, after which lateness spreads across successive jobs and dispatchers perform frequent overrides.","consequence":"Nominal utilization remains high while realized service reliability, schedule stability, and recovery capability deteriorate.","affected_objective":"Reduce expected lateness, overtime, disruption propagation, and rescheduling cost while preserving acceptable throughput and equitable service.","structural_mapping":[{"archetype_element":"Crowded field","domain_realization":"A nominal schedule packs jobs into almost all available capacity, leaving disruptions coupled across adjacent assignments.","claim_kind":"INFERENCE"},{"archetype_element":"Deliberate absence","domain_realization":"Selected capacity-time blocks remain intentionally unassigned to ordinary work as recovery reserve.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Protected empty region","domain_realization":"Constraints prevent routine dispatch from consuming reserve before a defined disruption or release condition occurs.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Positive form clarified by absence","domain_realization":"The reserve separates job blocks and localizes schedule deviation, making committed assignments more stable rather than merely reducing job count.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Meaningful empty state","domain_realization":"Operators can distinguish intentional idle reserve from no demand, missing data, equipment failure, or solver failure.","claim_kind":"HYPOTHESIS"},{"archetype_element":"Reintroduction","domain_realization":"Unused reserve returns to eligible work at a specified cutoff or is consumed by qualifying recovery work.","claim_kind":"HYPOTHESIS"}],"component_map":[{"component":"Omission Candidate","status":"adapted","domain_realization":"Candidate machine-, crew-, or vehicle-time blocks that may remain unassigned without violating hard commitments."},{"component":"Protected Empty Space","status":"adapted","domain_realization":"Explicit recovery-capacity blocks protected from routine assignment."},{"component":"Positive Form Relationship","status":"adapted","domain_realization":"Each reserve block is tied to the job chain or resource whose disruption propagation it is intended to interrupt."},{"component":"Attention Competition Map","status":"adapted","domain_realization":"A dependency and congestion map identifies jobs competing for the same capacity and the chains most exposed to propagated delay."},{"component":"Absence Boundary","status":"adapted","domain_realization":"Resource, duration, planning horizon, and eligibility constraints delimit each reserve block."},{"component":"Clarity or Effect Test","status":"adapted","domain_realization":"Historical replay tests whether protected reserve reduces propagation and realized cost relative to lost throughput."},{"component":"Rest and Pacing Zone","status":"adapted","domain_realization":"Recovery intervals are placed at high-risk transition points such as shift, route, setup, or job-chain boundaries."},{"component":"Meaning-of-Absence Check","status":"adapted","domain_realization":"State labels distinguish reserved idle capacity from outage, starvation, telemetry loss, infeasibility, and solver error."},{"component":"Reintroduction Trigger","status":"adapted","domain_realization":"A time cutoff or disruption threshold releases unused reserve to an eligible job queue."},{"component":"Accessibility and Recoverability Guardrail","status":"adapted","domain_realization":"Dispatchers can inspect reserve rationale, invoke an auditable override, and recover displaced jobs through the next reoptimization."},{"component":"Context Preservation Frame","status":"adapted","domain_realization":"Due dates, priorities, safety constraints, demand forecasts, and deferred-job records remain visible when capacity is left empty."}],"mechanism_dispositions":[{"slug":"architectural_void","disposition":"selected_load_bearing","contribution_type":"CORE_CAUSAL","adaptation_or_rejection":"Treat bounded unassigned capacity as a designed scheduling element whose surrounding commitments determine its location and value.","counterfactual_removal":"Without explicit reserve regions, the optimizer can refill all gaps and the proposed propagation-breaking lever disappears."},{"slug":"blank_or_rest_frame","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Adapt segment-boundary rest frames into short recovery intervals between vulnerable schedule segments.","counterfactual_removal":"Aggregate reserve could remain, but transition-specific deviations would be harder to absorb locally."},{"slug":"editorial_cut","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Removing jobs changes demand scope and priority policy; it is not required to reserve capacity.","counterfactual_removal":"No change; jobs remain queued rather than deleted."},{"slug":"empty_state_design","disposition":"selected_supporting","contribution_type":"SAFETY_GUARDRAIL","adaptation_or_rejection":"Label why capacity is idle and identify the authorized release action.","counterfactual_removal":"Operators may mistake reserve for failure or waste and prematurely fill it."},{"slug":"facilitation_silence","disposition":"incompatible","contribution_type":"NONE","adaptation_or_rejection":"The causal substrate is machine or service scheduling, not conversational participation.","counterfactual_removal":"No change."},{"slug":"focus_mode_or_control_hiding","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Hiding dispatch controls or alerts does not create capacity and could obstruct recovery.","counterfactual_removal":"No change to the scheduling lever."},{"slug":"margin_and_gutter_system","disposition":"selected_supporting","contribution_type":"OPERATIONAL","adaptation_or_rejection":"Encode reusable minimum buffer rules by resource and transition class so routine reoptimization preserves reserve.","counterfactual_removal":"Reserve could be manually specified, but consistency and resistance to gradual refill would weaken."},{"slug":"negative_space_logo","disposition":"incompatible","contribution_type":"NONE","adaptation_or_rejection":"Figure-ground symbolism has no relevant scheduling function.","counterfactual_removal":"No change."},{"slug":"pause_in_speech","disposition":"incompatible","contribution_type":"NONE","adaptation_or_rejection":"Live rhetorical pacing does not act on resource congestion.","counterfactual_removal":"No change."},{"slug":"sparse_layout","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Displaying fewer jobs may improve a dashboard but does not make the underlying schedule resilient.","counterfactual_removal":"No change to realized capacity or propagation."},{"slug":"whitespace","disposition":"considered_rejected","contribution_type":"NONE","adaptation_or_rejection":"Visual spacing may aid schedule inspection but is peripheral to the capacity intervention.","counterfactual_removal":"The reserve-capacity causal chain remains intact."}],"causal_chain":["[INFERENCE] Near-full nominal assignment couples adjacent jobs through shared finite capacity.","[HYPOTHESIS] Protected unassigned intervals give deviations a local absorption region.","[HYPOTHESIS] Boundary and release rules prevent routine work from consuming that region too early while returning genuinely excess reserve later.","[HYPOTHESIS] Fewer deviations propagate, reducing lateness, overrides, overtime, and rescheduling cost.","Throughput loss and idle-capacity cost are measured against those gains rather than presumed acceptable."],"baseline":"Ordinary deterministic rolling-horizon scheduling that optimizes nominal cost or utilization, with only incidental slack and dispatcher overrides after disruption.","nearest_rival":"Robust or stochastic scheduling over uncertainty scenarios without an explicit, operator-visible protected-reserve state and release rule.","authority_safety":{"affected_parties":["frontline workers whose workload or shifts may change","customers and downstream processes exposed to delay","dispatchers accountable for overrides","lower-priority claimants who could be deferred","owners of constrained equipment or service capacity"],"decision_authority":"The operations owner may approve a shadow test; any live reserve policy requires joint approval from scheduling, frontline operations, and the accountable safety or service owner.","authorized_first_step":"Run a preregistered historical replay on one resource pool and four representative weeks, comparing baseline, nearest rival, and protected-reserve variants on throughput, lateness, overtime, overrides, reserve use, and subgroup service levels; make no live dispatch changes.","excluded_actions":["Do not reserve capacity required for statutory, emergency, safety-critical, or minimum-service obligations.","Do not conceal deferred demand, missed service, uncertainty, or solver infeasibility.","Do not delete jobs or silently change priority classes.","Do not automate live release or override decisions during the first test."],"halt_rollback":"Reject the pilot variant if it violates any hard constraint or protected service level, materially worsens service for a claimant class, or loses more throughput than the preregistered bound without a compensating reduction in realized disruption cost; rollback is immediate because the first test is replay-only."}},"negative_tests":{"strongest_counterevidence":"The same replay benefit is matched or exceeded by the robust/stochastic rival, or observed disruption cost is explained by chronic undercapacity, invalid processing-time data, or infeasible commitments rather than propagation from packed schedules.","analogy_break":"Operational emptiness consumes scarce productive capacity and carries opportunity cost; unlike visual whitespace, it has no inherent clarifying effect and is useful only if quantified recovery benefits exceed lost output and distributional harms.","failure_condition":"Reserve is routinely filled early, placed away from vulnerable dependency chains, too small to absorb deviations, or so large that throughput and service losses dominate.","problem_falsifier":"Across representative records, high nominal utilization and small deviations do not predict propagated lateness, override frequency, overtime, or schedule instability after controlling for demand level and hard infeasibility.","intervention_falsifier":"With the problem present, explicit protected reserve plus release rules does not outperform baseline and nearest rival on preregistered realized-cost and service metrics at an acceptable throughput bound.","risks":["Idle capacity may worsen queues or defer lower-priority claimants.","Managers may treat visible reserve as waste and override it selectively.","Biased placement may shift reliability gains toward favored customers or resources.","State misclassification may preserve reserve during a real outage or release it during missing-data conditions.","Historical replay may underrepresent behavioral adaptation and rare disruptions."]},"null_rationale":null,"classification":{"candidate_kind":"MECHANISM_ADAPTATION","prior_art_status":"UNSEARCHED","evidence_maturity":"HYPOTHESIS"},"revision_change_log":{"revision_kind":"ORIGINAL","prior_problem_id":null,"prior_causal_lever_id":null,"problem_changed":false,"causal_lever_changed":false,"conceptual_changes":[],"operational_changes":[],"repairs_addressed":[]},"confidence":0.84,"generator_notes":"Closed-book structural inference. The candidate adapts designed absence into explicitly protected recovery capacity; it does not claim novelty or empirical effectiveness."}