{"schema_version":1,"research_id":"eoa_inverse_innovation_exp05_external_evaluation_20260803","source_assessment_id":"negative_space_design__astronomy_astrophysics:P1:v0","cell_id":"negative_space_design__astronomy_astrophysics","search_queries":["telescope scheduling target of opportunity interrupt queue dynamic scheduling transients observatory official","protected time reserve telescope schedule transient follow-up discretionary time","astronomical observatory queue scheduling ToO rapid response policy Gemini ESO official","telescope scheduler preemptible filler observations transient follow-up research","astronomy telescope \"reserved time\" transient follow-up schedule","observatory \"unallocated time\" target of opportunity telescope","Rubin target of opportunity reserved observing time transients policy","telescope schedule buffer time target of opportunity astronomy","site:gemini.edu/observing/phase-i/too Gemini Target Opportunity Rapid response queue","site:eso.org/sci/facilities/lasilla/sciops/doc \"Reserved Time\" nights DDT","site:eso.org/sci/observing/policies/too_policy.html RRM trigger interrupt ongoing observation","site:rubinobs.atlassian.net PSTN-056 target opportunity 3% observing time"],"sources":[{"source_id":"S1","title":"Review and Implementation of Rapid Response Observations on JWST","publisher":"Space Telescope Science Institute","url":"https://jwst-docs.stsci.edu/jwst-opportunities-and-policies/jwst-general-science-policies/review-and-implementation-of-rapid-response-observations-on-jwst","source_class":"OFFICIAL_GUIDANCE","publication_date":"Not stated","accessed_at":"2026-08-03","claims_supported":["JWST schedules are constructed weekly, with a 14-day preparation/execution interval during which inserting a transient can require schedule rebuilding.","Schedule changes generally affect multiple programs rather than producing a simple one-for-one swap.","Adding a transient observation requires replacing scheduled science, creating an explicit trade between science achieved and science deferred or lost.","JWST identifies TAC-reviewed ToO programs and Director-approved discretionary programs as authorization paths."]},{"source_id":"S2","title":"Dynamic Scheduling Guide","publisher":"National Radio Astronomy Observatory","url":"https://science.nrao.edu/facilities/vlba/observing/dynamicguide","source_class":"OFFICIAL_GUIDANCE","publication_date":"Not stated","accessed_at":"2026-08-03","claims_supported":["The VLBA maintains dynamic-scheduling eligibility for approved target-of-opportunity proposals.","Open time is filled from a dynamic queue, sometimes only hours before observing.","Selection accounts for proposal ratings, priority, array state, and weather, providing a close preemptible-filler comparator."]},{"source_id":"S3","title":"Special Scheduling Modes","publisher":"Las Cumbres Observatory","url":"https://lco.global/documentation/special-scheduling-modes/","source_class":"OFFICIAL_PRODUCT_DOCUMENTATION","publication_date":"Not stated","accessed_at":"2026-08-03","claims_supported":["LCO reruns a network-wide optimizer whenever a request arrives and reports typical rescheduling under ten minutes and request-to-start lag under fifteen minutes.","Rapid-response requests bypass normal scheduling and abort the observation in progress.","LCO reports a median rapid-response time of six minutes and limits the mode because it disrupts normal operation.","TAC-reviewed proposals and operational scheduling software provide identifiable authorization and implementation precedents."]},{"source_id":"S4","title":"The Rubin Observatory Target-of-Opportunity System in the First Year of Operations","publisher":"MacBride et al., arXiv","url":"https://arxiv.org/abs/2607.00217","source_class":"PRIMARY_RESEARCH","publication_date":"2026-06-30","accessed_at":"2026-08-03","claims_supported":["Rubin allocates 3% of total survey time to target-of-opportunity observations.","Rubin ToO observations use a distinct scheduling mode, interrupt other surveys, and require specialized workflows and communication channels.","Rubin trigger criteria are reviewed by the Survey Cadence Optimization Committee before adoption.","Rubin validates strategies through injected-event simulations and regards mock-alert exercises as essential before on-sky use."]},{"source_id":"S5","title":"AstroQ: Automated Scheduling of Cadenced Astronomical Observations","publisher":"Lubin et al., arXiv","url":"https://arxiv.org/abs/2506.08195","source_class":"PRIMARY_RESEARCH","publication_date":"2025-06-09","accessed_at":"2026-08-03","claims_supported":["AstroQ has been tested on Keck Observatory cadence programs.","It can optimize 3,680 observations of 200 targets over six months at five-minute resolution in about 120 seconds on a modern workstation.","The reported runtime supports technical feasibility of retrospective scenario generation and dynamic-rescheduling comparators."]},{"source_id":"S6","title":"Priorities for Visitor, Service, Technical, Reserved and Idle Nights","publisher":"European Southern Observatory","url":"https://www.eso.org/sci/facilities/lasilla/sciops/doc/sciOp_LSO-PRO-ESO-90000-2_techResTime.html","source_class":"OFFICIAL_GUIDANCE","publication_date":"2002-12-20","accessed_at":"2026-08-03","claims_supported":["ESO La Silla explicitly defines nights reserved for Director's Discretionary Time.","Reserved nights are protected for a specified contingent use and change status after an allocation decision.","If the DDT committee does not assign the time, reserved nights automatically convert to technical time.","This is direct prior art for protected non-assignment with an authorizer and release rule, although at whole-night rather than bounded within-night resolution."]},{"source_id":"S7","title":"Dynamic Scheduling: Target of Opportunity Observations of Gravitational Wave Events","publisher":"Almualla et al., arXiv","url":"https://arxiv.org/abs/2003.09718","source_class":"PRIMARY_RESEARCH","publication_date":"2020-03-21","accessed_at":"2026-08-03","claims_supported":["Dynamic rescheduling for transient follow-up has been implemented in gwemopt for the GROWTH and GRANDMA networks.","A simulation comparing a normally scheduled night with a night divided into blocks showed that failed observations could be rescheduled in the later block.","The work demonstrates block scheduling, prior-observation state tracking, and explicit before/after simulation as established comparators."]},{"source_id":"S8","title":"The Rapid Response Mode","publisher":"European Southern Observatory","url":"https://www.eso.org/sci/observing/phase2/SMSpecial/RRMObservation.html","source_class":"OFFICIAL_GUIDANCE","publication_date":"2024-08-21","accessed_at":"2026-08-03","claims_supported":["ESO's VLT Rapid Response Mode ends an ongoing observation and automatically executes a pre-prepared transient observation block after a valid trigger.","Programs require prior approval and Phase 2 preparation, and only designated activators can use the mechanism.","Rapid-response observations override ordinary observations subject to protected-status and operational exceptions.","ESO preserves operator, instrument, observability, time-allocation, and misuse guardrails relevant to authority and safety feasibility."]}],"problem_evidence":{"support":"STRONG","rationale":"The problem is visible in official operations documentation: JWST states that inserting a transient during its committed scheduling horizon requires rebuilding the schedule, generally affects multiple programs, and replaces already scheduled science. LCO, ESO, and Rubin independently document interruption-based responses and their operational costs. These sources establish that rapid transient follow-up competes with committed observations and matters scientifically and operationally. They do not establish how often protected vacancy would outperform existing procedures at any particular facility.","source_ids":["S1","S3","S4","S8"]},"stakeholder_evidence":{"support":"STRONG","rationale":"Identifiable adopters and authorizers include observatory scheduling and operations groups, telescope-allocation committees, observatory directors, ESO's OPC/DDT authorities, STScI's TAC and Director, and Rubin's Survey Cadence Optimization Committee. Their published policies express demand for rapid transient response, controlled disruption, approved trigger criteria, and schedule-impact management. No source expresses demand specifically for within-night protected vacancy.","source_ids":["S1","S3","S4","S6","S8"]},"prior_art":{"proximity":"SUBSTANTIAL_COLLISION","closest_analogues":[{"name":"ESO La Silla reserved nights for Director's Discretionary Time","similarity":"Directly treats deliberately unassigned observing time as a protected resource, identifies an allocation authority, and specifies a release/conversion rule.","remaining_difference":"The documented unit is a whole night reserved broadly for DDT and released to technical time, not a short response window centered on a forecast information-release point with restoration from a recoverable science side queue.","source_ids":["S6"]},{"name":"VLBA open dynamic time filled from an eligible queue","similarity":"Maintains open time, selects eligible proposals close to execution, and uses priority, feasibility, array-state, and weather information.","remaining_difference":"The open interval is intended to be filled from the queue rather than protected as genuine vacancy until a prespecified transient decision deadline.","source_ids":["S2"]},{"name":"ESO and LCO rapid-response interruption modes","similarity":"Provide approved transient triggers, rapid operator workflows, priority rules, and immediate access to telescope capacity.","remaining_difference":"They obtain capacity by terminating or aborting incumbent observations rather than by preserving an empty interval in advance.","source_ids":["S3","S8"]},{"name":"Rubin dedicated ToO allocation and specialized scheduler mode","similarity":"Preallocates a bounded share of observing time to transient response, uses approved trigger classes, interrupts ordinary survey observations, and validates behavior through simulation.","remaining_difference":"The 3% allocation is an accounting envelope, not evidence that particular forecast decision windows remain genuinely unassigned before a trigger.","source_ids":["S4"]},{"name":"AstroQ and gwemopt dynamic rescheduling","similarity":"Demonstrate computationally tractable schedule recomputation, block scheduling, state-aware rescheduling, and simulation-based comparisons.","remaining_difference":"They optimize assigned observations after changing information; they do not test whether protected non-assignment itself changes response performance relative to equally authorized preemptible filler.","source_ids":["S5","S7"]}],"distinctive_claim_remaining":"For a prespecified transient-information decision point, holding a short interval genuinely vacant until a release deadline—rather than filling it with a preemptible observation, interrupting an incumbent, or dynamically rebuilding the queue—will increase deadline-feasible follow-up or reduce churn and failed restoration enough to outweigh expired-window and displaced-science costs. This contrast is falsifiable, but no reviewed source supplies comparative evidence for it.","confidence":"HIGH"},"implementation_evidence":{"support":"MODERATE","rationale":"Existing observatories demonstrate trigger authorization, automated interruption, dynamic queue selection, minute-scale recomputation, simulations, and protected-time governance. These make a read-only replay technically and procedurally credible. Feasibility remains facility-specific because the proposal needs reconstructable event timestamps, visibility and instrument states, setup dependencies, displaced-program records, scheduler participation, and approval to interpret science tradeoffs. No source validates the proposed display semantics or restoration rule.","source_ids":["S1","S2","S3","S4","S5","S6","S7","S8"]},"scores":{"meaningful_impact":{"score":4,"rationale":"Avoiding missed, time-sensitive transient observations while reducing disruption could materially improve scientific yield and fairness, although prevalence and effect size are unmeasured.","source_ids":["S1","S4"]},"stakeholder_pull":{"score":4,"rationale":"Multiple observatories operate formal rapid-response programs with named authorizers and restricted allocations, demonstrating strong pull for the underlying response problem but not specifically for vacancy.","source_ids":["S1","S3","S4","S8"]},"incremental_advantage":{"score":2,"rationale":"The vacancy-versus-filler advantage is plausible but unsupported; fast dynamic scheduling and interruption already achieve minute-scale responses, and reserved observing time is established.","source_ids":["S2","S3","S5","S6","S8"]},"distinctiveness_plausibility":{"score":2,"rationale":"The exact within-night, forecast-centered, recoverable-side-queue configuration was not found, but ESO's protected reserved nights substantially collide with the defining non-assignment mechanism.","source_ids":["S6"]},"technical_implementability":{"score":4,"rationale":"Existing schedulers recompute thousands of observations quickly, and Rubin and gwemopt demonstrate simulation-based transient scheduling tests. Integration and data reconstruction remain local challenges.","source_ids":["S4","S5","S7"]},"adoption_authority_feasibility":{"score":3,"rationale":"Allocation committees, directors, and operations groups already authorize ToO modes and protected time, but a live vacancy trial would redistribute scarce approved science and therefore needs formal local approval.","source_ids":["S1","S4","S6","S8"]},"evidence_readiness":{"score":3,"rationale":"A frozen-log replay is bounded and low-risk, but decisive evidence requires proprietary operational records, local dependency knowledge, and scheduler judgments not available on the open web.","source_ids":["S1","S4","S7"]},"safety_net_benefit":{"score":3,"rationale":"An expiry rule and recoverable side queue could reduce irreversible displacement, but hidden setup and calibration dependencies could make apparent restoration unreliable.","source_ids":["S1","S2","S8"]},"scalability":{"score":3,"rationale":"The pattern could transfer across queue-scheduled facilities, but trigger classes, weather, instruments, cadence obligations, and allocation governance differ enough to require facility-specific calibration.","source_ids":["S2","S3","S4","S6"]}},"score_confidence":"MODERATE","costs":{"first_evidence":{"band_2026_usd":"10K_TO_50K","scope":"One retrospective shadow study covering at most 12 completed observing periods, including data extraction, reconstruction, prespecified scenario generation, two scheduler reviews, and analysis.","confidence":"LOW","assumptions":["Approximately 120-300 loaded staff hours across an analyst, scheduler, and transient-science reviewer.","Existing logs and scheduler or a faithful simulator are available without major software procurement.","No telescope time or live-operations change is included.","No direct public labor-rate source was found."],"source_ids":["S4","S5","S7"]},"initial_deployment_startup":{"band_2026_usd":"50K_TO_250K","scope":"Facility-specific implementation of window state, authorization and expiry rules, recoverable side queue, operator display, audit logging, integration tests, and governance review.","confidence":"LOW","assumptions":["Roughly 0.5-1.5 FTE-equivalent over three to six months across scheduling software, operations, UX, and policy staff.","Existing scheduler and alert infrastructure are extended rather than replaced.","Instrument-control or flight-software certification is not required.","No public source prices this exact implementation."],"source_ids":["S3","S4","S5","S8"]},"operational_launch":{"band_2026_usd":"50K_TO_250K","scope":"A three-to-six-month prospective shadow or tightly limited operational trial, including training, monitoring, incident review, investigator communication, and independent science-displacement review.","confidence":"LOW","assumptions":["Launch begins in shadow mode and only later uses a small number of approved windows.","Existing operations staff can absorb most shifts, with dedicated engineering and evaluation support.","Any reserved telescope time is recovered through the eligible side queue when no trigger occurs.","Opportunity cost of lost science could move the total above this band if restoration performs poorly."],"source_ids":["S1","S3","S4","S8"]},"annual_recurring":{"band_2026_usd":"50K_TO_250K","scope":"Ongoing rule maintenance, trigger review, scheduler support, audit and outcome analysis, operator training, and residual resource-equivalent cost of windows that expire or fail restoration.","confidence":"LOW","assumptions":["Approximately 0.25-1.0 FTE-equivalent recurring support.","Protected windows cover only a small fraction of observing time.","Most unused windows are successfully returned to approved observations.","The resource-equivalent value of scarce telescope time is not publicly established in the reviewed sources and could dominate the band."],"source_ids":["S3","S4","S6","S8"]}},"verified_pipeline_gates":{"externally_supported_problem":{"status":"YES","reason":"Official JWST documentation directly reports schedule rebuilding, multi-program effects, and replacement of scheduled science for transient insertions; other observatories document disruptive interruption.","source_ids":["S1","S3","S8"]},"externally_credible_adopter_or_authorizer":{"status":"YES","reason":"TACs, observatory directors, operations groups, ESO's allocation authorities, and Rubin's SCOC are identifiable bodies already authorizing rapid-response or reserved-time practices.","source_ids":["S1","S4","S6","S8"]},"distinct_testable_incremental_claim":{"status":"YES","reason":"Protected genuine vacancy can be compared directly with interruption, dynamic rescheduling, and preemptible filler on deadline-feasible opportunities, churn, displacement, restoration, and unused time.","source_ids":["S2","S3","S5","S6","S7","S8"]},"bounded_next_evidence_step":{"status":"YES","reason":"A maximum-12-period frozen-log replay with prespecified comparators, parameter grid, outcomes, and falsifiers is bounded and does not change live allocations.","source_ids":["S4","S5","S7"]},"no_unresolved_safety_or_authority_stop":{"status":"YES","reason":"The first step is read-only; live scheduling, engineering, calibration, and safety decisions remain excluded and existing policies demonstrate workable approval and operator-override structures.","source_ids":["S1","S4","S8"]},"credible_cost_scope_and_range":{"status":"UNCERTAIN","reason":"Work packages and broad staffing assumptions can be bounded, but no reviewed source supplies local loaded labor rates or the resource-equivalent value of displaced telescope time, which may dominate recurring cost.","source_ids":["S1","S4","S6"]}},"next_evidence_step":"Partner with one queue-scheduled observatory and preregister a retrospective replay of no more than 12 completed transient periods with reconstructable schedules, alert/localization updates, weather and instrument state, visibility constraints, setup dependencies, and displacement outcomes. Freeze information at each historical decision time. Compare (A) actual or fully assigned baseline, (B) existing interruption policy, (C) continuous dynamic rescheduling, (D) preemptible filler, and (E) protected genuine vacancy over a small prespecified grid of window placements and durations. Measure deadline-feasible follow-up, response latency, schedule churn, incumbent time lost, restoration success, expired-window time, operator state misclassification, and science-priority adjudication. Falsify the problem if failures primarily reflect weather, visibility, instrument capability, or decision latency rather than schedule occupancy. Falsify the intervention if protected vacancy does not improve deadline-feasible follow-up or churn over the best comparator, or if authorized reviewers judge its displaced/expired science cost unacceptable.","blocking_evidence":["Facility-level prevalence of missed or disruptive transient follow-up attributable specifically to schedule occupancy rather than weather, visibility, instrument capability, or decision latency.","Comparative frozen-log results showing whether genuine vacancy beats a preemptible filler program, interruption, and modern dynamic rescheduling.","Reconstructable setup, cadence, calibration, and restoration dependencies for observations moved into the side queue.","Operator evidence that the protected-window state is correctly interpreted and not confused with outage, incomplete scheduling, or ordinary slack.","Local authorization criteria for eligible event classes, activation authority, conflict resolution, and acceptable science displacement.","Loaded labor rates and the local resource-equivalent value of telescope time needed to validate cost bands."],"research_disposition":"PARTNERED_RESEARCH_PROGRAM","world_novelty_boundary":"The search found substantial prior art for protected reserved observing time, dynamic open-time filling, dedicated ToO allocations, interruption, and rapid rescheduling. It did not establish whether the exact combination of a short forecast-centered genuine vacancy, explicit empty-state semantics, expiry into a recoverable science side queue, and comparator-based validation has previously been implemented. World novelty, patentability, freedom to operate, market size, and realized impact remain unmeasured.","arm":"COMPLETE_PROPOSAL_PORTFOLIO","candidate_version":0,"controller_recommendation":{"action":"STOP_EMPIRICAL_RESEARCH_NEEDED","repairable":false,"material_progress_observed":false,"progress_targets":["Secure one observatory partner with authority to provide frozen operational logs and scheduler participation.","Estimate occupancy-attributable problem prevalence from the bounded sample.","Run the preregistered five-arm replay and report all displacement and restoration costs.","Demonstrate a prespecified advantage over preemptible filler and dynamic rescheduling, not merely over a static fully assigned schedule.","Validate operator interpretation of the reserved state and document authority, safety, calibration, and rollback guardrails.","Replace assumption-based cost bands with facility-specific loaded labor and telescope-time resource equivalents."],"reason":"Open sources verify the problem, credible authorizers, technical feasibility, and substantial collision with existing reserved-time and rapid-response practices. The decisive remaining claim—whether genuine vacancy near a forecast decision point outperforms filler, interruption, and dynamic rescheduling—requires proprietary logs, scheduler participation, and comparative replay or live testing. That evidence cannot be resolved by further bounded web research."},"proposal_index":1}