{"id":"db8009c0-5760-4f8c-a8ef-dd233b9a0861","arxiv_id":"2509.03005","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":5,"one_line_summary":"The SSS framework maps the sky to HEALPix regions, assigns adaptive weights to targets, and schedules multi-telescope observations; a 12-telescope run produced 41,319 observation arcs.","lead":"This paper describes a software framework, the Sky Survey Schedule (SSS), that coordinates dozens of telescopes for observing space objects, monitoring asteroids, and following up cosmic transients. A 12-telescope demonstration produced a schedule covering thousands of satellites, but the core optimization algorithm and comparisons to existing schedulers are not presented.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central multi-site claim is unsupported: the only demonstration uses 12 telescopes at one station, and §5 defers multi-site collaboration to future work; the scheduler's optimization algorithm is also unspecified.","rationale":"The reader's weakest assumption focuses on the unspecified optimization algorithm in §3.3. That concern is real, but it is secondary to a more direct claim–evidence gap: the paper's headline contribution is multi-site network coordination, yet the only demonstration uses a single station, and the future-work section explicitly says multi-site collaboration remains to be done. This is not a matter of hidden algorithm details; the paper itself admits the capability is not present. The demonstration's 2,488 targets and 41,319 arcs all arise from 12 telescopes at one longitude/latitude, so they cannot validate the abstract's 'multi-site observational networks' claim. A CONDITIONAL verdict remains appropriate: the framework may be useful for single-station scheduling, but the authors must either remove or soften the multi-site claim, or provide a multi-site demonstration. The optimization algorithm also needs disclosure, but the first-order issue is the scope mismatch. I therefore keep the reader's CONDITIONAL verdict, while shifting the primary emphasis from the black-box optimizer to the unsubstantiated multi-site capability.","tokens_in":9543,"tokens_out":5001,"duration_ms":54806,"concrete_test":"Run the scheduler on a simulated network of at least two well-separated geographic sites (e.g., Yao'an at E101° and a second station at E70°), using the same catalog and task set. Add the second site to the Station database, submit a shared Key Task visible from both sites, and inspect the generated Observation Schedule: a true multi-site scheduler should either assign the task to one site or coordinate a joint observation; two independent single-station plans would falsify the abstract's multi-site claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract states SSS 'facilitates coordinated operations across multi-site observational networks comprising dozens of instruments.' The sole demonstration (§4) uses 12 telescopes at Yao'an Observatory—one geographic site. §3.3 describes optimizing 'task allocation among multiple instruments within the single station,' not across sites. §5 explicitly lists 'expanding the framework to support multi-site collaborative observation and data fusion' as future work, confirming that multi-site coordination is not implemented. The central claim of the paper—a network-level scheduler—is therefore not supported by the presented evidence; even if the optimization algorithm were fully specified, it would only address a single-station problem. This is a direct claim–evidence gap, not a matter of tuning. A secondary issue: the scheduling step in §3.3 ('selects the appropriate sky region based on two-dimensional weights') is never defined as an objective function or algorithm, so the asserted 'optimization' is not reproducible.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper describes the Sky Survey Schedule (SSS) framework, a software system for automated scheduling of optical space-object observations. After introducing the need for coordinated, adaptive scheduling in space surveillance, the authors describe a HEALPix-based division of the sky into latitude bands and sectors, and penalty functions for solar phase angle, moonlight, and elevation. They then present a target-weighting scheme combining base weights, reliability factors, time-recovery terms, and operator-defined priorities, with dynamic feedback that resets weights after observation, recovers weights over time, and escalates anomalies. The mission scheduler combines these weights with visibility constraints to produce observation plans. A demonstration uses 12 telescopes at Yao'an Observatory with a catalog of 4,164 MEO/HEO objects, producing 2,488 scheduled targets and 41,319 observation arcs, and reporting revisit behavior for three high-priority BeiDou targets. The paper concludes with future work items including multi-site collaboration and quantitative evaluation.","tokens_in":9817,"tokens_out":2642,"duration_ms":32529,"significance":"The practical problem addressed by SSS is real and current: optical space-surveillance networks need schedulers that can allocate limited telescope time across many targets while responding to maneuvers, weather, and equipment failures. The proposed architecture has sensible components: HEALPix spatial indexing for fast region lookup, penalty weights for environmental conditions, and a closed-loop weight-update mechanism for prioritization. If fully implemented and validated, such a framework could be useful for observatory operations. The paper also provides concrete numbers from a worked example, which is a useful starting point. However, the significance is currently limited by two major gaps: the network-level claim is not supported by the presented single-site demonstration, and the scheduler's optimization step is described only qualitatively. As presented, the paper is an architecture and concept description rather than a validated scheduling solution.","major_comments":[{"comment":"The abstract claims SSS 'facilitates coordinated operations across multi-site observational networks comprising dozens of instruments,' but the only demonstration (§4) uses 12 telescopes at one geographic site (Yao'an Observatory). Section 3.3 itself states that the scheduler 'optimizes task allocation among multiple instruments within the single station,' and §5 explicitly lists 'expanding the framework to support multi-site collaborative observation and data fusion' as future work. The evidence therefore does not support the network-level claim. This is a direct claim–evidence gap that affects the paper's central message. The authors should either present a multi-site demonstration or substantially revise the abstract and introduction to present SSS as a single-site scheduler with a multi-site extension as planned work.","section":"Abstract and §4/§5"},{"comment":"The core scheduling step is not specified in a reproducible way. The text says the scheduler 'selects the appropriate sky region based on two-dimensional weights' and 'optimizes task allocation among multiple instruments,' but no objective function, formal constraint set, or optimization algorithm is given. It is unclear whether this is a greedy highest-weight selection, a heuristic local search, or a global optimization. Because the demonstration's success (including the stated 2,488 targets and 41,319 arcs) depends on this unspecified selection process, the asserted 'optimization' is not reproducible. Please provide a mathematical formulation of the scheduling problem and the algorithm used, or explicitly state that the current implementation is a heuristic and characterize its behavior.","section":"§3.3 Mission scheduler"},{"comment":"The demonstration is descriptive and lacks any quantitative evaluation of schedule quality. There is no baseline or comparison against alternative scheduling strategies (e.g., random, greedy, or static-priority scheduling), no metric for schedule optimality or resource utilization, and no error bars or run-to-run variability. The statement that 'the detailed observation statistics validate its capacity to ensure both coverage completeness and priority target responsiveness' is not supported by the reported numbers alone. In particular, the frequent observation of priority targets is partly a direct consequence of assigning them high weights in the scheduler, not evidence that the scheduler achieves globally good coverage or conflict-free assignment. Please add quantitative evaluation with appropriate comparisons and metrics.","section":"§4 Operation and Discussion"},{"comment":"Equation (7) is garbled in the manuscript: the set-builder notation and the '9' fragment make the formal definition of the valid target set unreadable. Since this equation defines the candidate target set that feeds the scheduler, it is load-bearing for the framework description. Please rewrite it cleanly and ensure the conditions C_elev, C_night, C_eclipse, C_arc, and C_mag are each defined without OCR-like artifacts.","section":"Equation (7) and §2.2"}],"minor_comments":[{"comment":"Typo: 'Th is framework' should be 'This framework'.","section":"Abstract"},{"comment":"The formula for the solar phase angle penalty is unclear: '1 cos α' is missing an operator and it is not obvious how the weight decreases with α. Please clarify the expression.","section":"Equation (3)"},{"comment":"The moonlight penalty uses values (0, 0.3, 1) with thresholds φ and 2φ, but the text says 'these parameter values are preliminary and will be updated upon completion of a detailed follow-up study.' Given this admission, it may be appropriate to frame the current values as illustrative rather than definitive.","section":"§2.1/Equation (4)"},{"comment":"The red curve is labeled 'cumulative coverage percentage,' but the axis is not labeled and the denominator for the percentage is not stated. Please specify what 'coverage' means relative to (e.g., all 3,040 visible targets) and add axis labels.","section":"Figure 8"},{"comment":"The reference list is sparse relative to the breadth of the claimed framework; in particular, no reference is given for existing scheduling methods beyond a single multi-objective optimization paper. Adding a short comparison with previous sensor tasking and scheduling approaches would help position SSS.","section":"General presentation"}],"recommendation":"major_revision","confidential_remarks":"This is a conference paper whose strength is an architecture description with a concrete but limited demonstration. The main revision should focus on aligning claims with evidence: either add a multi-site demonstration or scale back the abstract and introduction. The unspecified scheduler is the most serious technical issue; even a clear statement that the current implementation uses a greedy region-selection heuristic with conflict resolution would be an improvement. I would be willing to look at a revised version with these points addressed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a clear engineering write-up of a real scheduling system, but the paper's flagship claim—coordinating multi-site networks of dozens of instruments—is not what the demonstration shows. The only demonstration is 12 telescopes at a single station, and Section 5 says multi-site collaboration is future work. The scheduling core is also underspecified. If you read it as a status report on a single-station scheduler, it's plausible. If you read the abstract literally, the evidence isn't there.\n\nWhat's genuinely useful: the modular architecture (ephemeris, weight adjustment, mission scheduler) is laid out in enough detail to follow; HEALPix sky-region mapping with configurable overlap is a reasonable choice; the adaptive weight feedback (decay, recovery, anomaly escalation) is clearly described; the 12-telescope run produced 41k arcs over 2,488 targets and the three priority targets behaved as intended. That is a legitimate engineering demonstration, and the system appears operational.\n\nWhere it is soft:\n- Multi-site claim. The stress-test note is right. Section 3.3 says the scheduler optimizes allocation 'within the single station'; Section 4 uses one site; Section 5 defers multi-site. Replace 'multi-site' in the abstract with 'multi-instrument single-station' or add at least one genuinely multi-site run.\n- Scheduler black box. 'Selects the appropriate sky region based on two-dimensional weights' and 'optimizes task allocation' is not an algorithm. No objective function, no constraint model, no optimality guarantee, no greedy vs global discussion. This is the load-bearing part of the framework; it needs specification.\n- No quantitative evaluation. No baseline, no comparison to greedy/round-robin, no runtime, no optimality metric. The observation counts for C7/C10/C56 are exactly what you'd expect from setting their weights high—so they show the feedback loop works, but they don't test overall schedule quality.\n- No artifacts. No code, no configs, no target catalog. Not required for an IAC paper, but it caps reproducibility.\n- Minor: the 'probabilistic resource allocation strategy' mentioned in Conclusions never appears elsewhere.\n\nMy take: the reader's condition is fair, but slightly generous. The system may work; the paper just doesn't let us check the core. I would send it to peer review with a request for a major revision that specifies the scheduler, reports a baseline comparison, and either drops or backs the multi-site language. For a reading group, it's useful as an example of claim–evidence mismatch in otherwise competent systems papers.","headline":"A plausible single-station scheduler is described clearly, but the paper claims multi-site network coordination while only demonstrating one station and leaving the actual optimization algorithm unspecified.","tokens_in":10219,"tokens_out":2099,"would_cite":false,"duration_ms":24712,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A unified sky-weight matrix turns multi-telescope networks into self-adjusting schedulers for space-object monitoring.","keywords":["space situational awareness","optical observation scheduling","HEALPix","multi-telescope coordination","adaptive weighting","space debris cataloging","time-domain astronomy"],"falsifier":"Re-run the Section 4 demonstration with the optimizer disabled (pick the single highest-weight region at each time step, ignoring cross-instrument conflict) and compare cumulative coverage and priority-target revisit statistics; if the plan is unchanged or barely worse, the optimization step adds nothing beyond the weighting bookkeeping.","tokens_in":9451,"feed_emoji":"🔭","tokens_out":6671,"duration_ms":76021,"temperature":0.7,"pith_summary":"This paper argues that telescope-network scheduling can be reduced to a single bookkeeping structure: every sky region and target carries a time-dependent weight, and the scheduler repeatedly observes the highest-weight regions while feeding results back into the weights. The Sky Survey Schedule (SSS) framework combines environmental penalty functions (solar phase angle, moonlight, elevation) with dynamically updated target priorities, so stable objects fade, long-unobserved objects recover, and newly detected or lost objects jump to the front of the queue. In a demonstration across 12 telescopes with a catalog of 4,164 medium- and high-Earth-orbit objects, the scheduler produced 2,488 scheduled targets and 41,319 observation arcs, while keeping priority targets on their assigned revisit modes. If the framework works as described, it would give space-surveillance networks a practical path to closed-loop, near-real-time scheduling under weather and equipment failures.","feed_headline":"Sky-weight matrix drives 41,319 telescope observations","feed_subtitle":"Closed-loop weights re-prioritize lost or anomalous space objects while routine survey coverage continues.","key_machinery":"The load-bearing object is the two-dimensional (sky-region by time) weight matrix built on HEALPix, an equal-area, hierarchical pixelization of the sphere. HEALPix gives each cell uniform area and fast hierarchical retrieval, so targets, instruments with different fields of view, and search grids can be cross-referenced quickly. The product-form penalty weights (solar phase times moonlight times elevation) convert observational constraints into comparable numbers, and the dynamic target-weight update loop—reset on success, recovery on neglect, escalation on anomaly—makes the schedule self-adjusting. Together they turn scheduling into a matter of selecting the region with the highest combined","core_discovery":"The central claim is that coordinated scheduling over dozens of instruments can be achieved by mapping targets and environmental conditions onto a common space-time priority matrix. Each sky region is assigned a comprehensive weight as the product of penalty factors for solar phase angle, moonlight, and low elevation; each target carries a base weight that is decayed after successful observation, recovered as it goes unobserved, and escalated for anomalies. The two families of weights are combined into two-dimensional priority values, one per sky region per time step, and the Mission Scheduler selects regions and allocates instruments from these values. The demonstration of the framework use","pith_inferences":["A reader could test whether the optimizer is load-bearing by running the same demonstration with the selection step replaced by a rule that simply picks the top-weighted region at each time step; if cumulative coverage and priority-target statistics are unchanged, the value of the framework is in the weighting and bookkeeping, not in the optimization.","The same two-dimensional weighting architecture could be lifted to other alert-driven surveys, such as gravitational-wave or neutrino follow-up, by swapping the orbital-catalog priorities for external alert maps; the paper does not mention this application.","The demonstration fixes the field of view at 7 by 7 degrees and the overlap at 1 degree, meaning the sky partition is tuned to this particular station; scaling to heterogeneous networks will require solving for region size and overlap assignment rather than presetting them."],"forward_implications":["Multi-site networks could run off one shared priority matrix, so telescopes with different fields of view and sky-division schemes can be coordinated via HEALPix cell lookup rather than bespoke per-site scheduling.","Anomaly response becomes automatic: targets lost to maneuvers get local search grids, unmatched tracks get feasible orbital regions, and both enter the Key Task database with elevated weights.","Catalog maintenance improves through a feedback loop: successful observations reset weights, neglected targets recover, and long-term coverage is driven by actual observation history.","Scheduled plans can absorb weather and equipment failures dynamically, because missing results feed back and trigger re-scheduling in the next cycle.","Priority targets can be given guaranteed revisit cadence without separating the survey and tracking pipelines, as demonstrated by the six observations of one target at roughly 1.5-hour intervals and fifteen observations of another at roughly 29-minute intervals."],"supporting_citations":[{"why":"Supplies the HEALPix equal-area hierarchical pixelization used to index sky regions and retrieve targets.","marker":"[5]"},{"why":"Defines the large-scale multi-objective sensor-tasking problem that SSS is positioned to address.","marker":"[4]"},{"why":"Provides short-arc orbit-determination methods used to construct feasible orbital regions for unmatched observations.","marker":"[7]"},{"why":"Supports the preliminary lunar-avoidance weight values, which the paper says will be refined by this lunar-impact study.","marker":"[6]"},{"why":"Makes the case for wide-field optical telescopes as the basis for the observation networks SSS schedules.","marker":"[3]"}],"fun_headline_variants":["Closed-loop weights steer 41,319 telescope observations","Adaptive weights re-prioritize lost and anomalous space objects","SSS framework: one matrix schedules dozens of telescopes","Weight matrix prioritizes anomalies across multi-site networks","Feedback loop boosts lost objects, keeps survey running"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The whole demonstration depends on the Mission Scheduler actually choosing good regions and avoiding conflicts; the paper describes 'selects the appropriate sky region based on two-dimensional weights' without specifying the objective function or the optimization algorithm, so if selection is only locally greedy, the near-real-time global optimization claim is unsupported.","fun_headline_variants_meta":{"raw":{"variants":["Closed-loop weights steer 41,319 telescope observations","Adaptive weights re-prioritize lost and anomalous space objects","SSS framework: one matrix schedules dozens of telescopes","Weight matrix prioritizes anomalies across multi-site networks","Feedback loop boosts lost objects, keeps survey running"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000706,"raw_usage":{"total_tokens":3027,"prompt_tokens":758,"completion_tokens":2269,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":502,"completion_tokens_details":{"reasoning_tokens":2192}},"tokens_in":502,"tokens_out":2269,"duration_ms":16826,"temperature":1.0,"reasoning_tokens":2192,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T11:12:29.847595+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Re-run the Section 4 demonstration with the optimizer disabled (pick the single highest-weight region at each time step, ignoring cross-instrument conflict) and compare cumulative coverage and priority-target revisit statistics; if the plan is unchanged or barely worse, the optimization step adds nothing beyond the weighting bookkeeping.","supporting_citations":[{"cited_title":"HEALPix: A framework for high - resolution discretization and fast analysis of data distributed on the sphere,","cited_arxiv_id":null,"evidence_quote":"Supplies the HEALPix equal-area hierarchical pixelization used to index sky regions and retrieve targets."},{"cited_title":"Efficient high -dimensional multi -objective optimisation method for large -scale sensor tasking,","cited_arxiv_id":null,"evidence_quote":"Defines the large-scale multi-objective sensor-tasking problem that SSS is positioned to address."},{"cited_title":"Orbit determination with very short arcs: II. Identifications,","cited_arxiv_id":null,"evidence_quote":"Provides short-arc orbit-determination methods used to construct feasible orbital regions for unmatched observations."},{"cited_title":"Quantifying the Lunar Imp act on Astronomical Observations Using a Machine Learning Framework for Visibility Prediction,","cited_arxiv_id":null,"evidence_quote":"Supports the preliminary lunar-avoidance weight values, which the paper says will be refined by this lunar-impact study."},{"cited_title":"Optical design study of the wide field survey telescope (WFST),","cited_arxiv_id":null,"evidence_quote":"Makes the case for wide-field optical telescopes as the basis for the observation networks SSS schedules."}],"review_version":1}