{"id":"1a8971a9-02d8-47ee-84df-7f5c77e4ce5d","arxiv_id":"2607.10888","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Multi-objective dynamic programming produces globally optimal Pareto fronts balancing ΔV and exoplanet yield for LEO nulling-interferometer observation schedules under realistic ConOps and observability constraints.","lead":"This paper builds a multi-objective dynamic-programming scheduler that trades fuel use against expected exoplanet detections for low-Earth-orbit formation-flying nulling interferometers. Mission designers get a global Pareto policy that respects realistic visibility, ConOps, and transfer constraints.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection beyond the reader's already-flagged fuel-model assumption; the global-optimality claim is the only soft spot that still needs a concrete check.","rationale":"The paper’s contribution is a carefully specialized multi-objective DP scheduler for LEO linear-formation nulling interferometers. The mathematics is standard, the ConOps and observability constraints are faithfully encoded, and the numerical comparisons across terminator vs. noon-midnight families are internally consistent. Absolute yield and ΔV figures rest on the empirical station-keeping model the reader already flagged; that is a calibration uncertainty, not an internal inconsistency. The only additional technical softness is the admitted loss of strict optimality when duplicates are pruned. Because the authors themselves note the issue and because a simple re-run of the existing algorithm can quantify its practical impact, the concern does not move the verdict. CONDITIONAL remains appropriate; no stronger rejection is warranted and no upgrade to ACCEPT is justified without released code or a verified revisit-free front. Confidence stays high: the text is fully readable and the engineering utility is clear.","tokens_in":18444,"tokens_out":531,"duration_ms":6264,"concrete_test":"Re-run Alg. 1 for Case 1.1b both with and without the duplicate-removal step (or with an explicit visited-set state that restores Markovianity). If any of the three highlighted paths on the Pareto front of Fig. 7a changes its sequence of stars or its (yield, ΔV) coordinates by more than ~10 %, the “globally optimal” language must be qualified; otherwise the claim stands under the paper’s own models.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest assumption (analytical transfer + empirical station-keeping coefficients in Eqs. 17–21) is correctly identified as the main source of absolute-number uncertainty, but it does not undermine the paper's central engineering claim: that a multi-objective DP produces a usable Pareto-front policy under the stated ConOps and observability models. The only remaining load-bearing softness is the authors' own admission that duplicate-removal (Alg. 1, after Eq. 32) breaks the Markov property and therefore strict global optimality of the returned front. Because the 6-month results are presented as “globally optimal,” any material change in the front once revisits are properly forbidden would weaken the strongest claim. All other modeling choices (visibility fractions, yield Monte-Carlo, frozen-orbit reduction) are either standard or explicitly optional and do not threaten the comparative conclusions across orbit families.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The manuscript develops a multi-objective dynamic-programming scheduler for LEO linear-formation nulling-interferometry missions. It encodes Sun exclusion, Earth/solar occultation, baseline oscillation, optional frozen-ROE (δix≈0) constraints, operational dwell, transfer ΔV (via ROE bounds with a standby intermediate), and an empirical station-keeping cost into a reachable-set Bellman update that returns a Pareto front of cumulative yield versus propellant. Yields are precomputed from the HWO stellar catalog with a simplified Bracewell SNR and Monte-Carlo completeness. Numerical results for six Sun-synchronous orbits over a 6-month horizon show expected detections of order ~4–10 for ΔV budgets of a few to ~10 m/s, and compare terminator versus noon–midnight families with and without the frozen-orbit constraint.","tokens_in":18737,"tokens_out":1235,"duration_ms":20227,"significance":"If the modeling assumptions hold, the work is a useful and timely contribution: it is among the first multi-objective DP applications to distributed space-telescope scheduling, produces a reusable policy (not a single open-loop path) that can restart after revisits or fuel changes, and gives concrete comparative guidance across absolute-orbit families under a realistic ConOps. Strengths include the clean specialization of the multi-objective Bellman update (Eq. 32), the ROE-based transfer geometry with standby relaxation (Eqs. 17–20), and the systematic orbit-family comparison (Figs. 8–10). Absolute science-return numbers remain model-dependent, but the relative ranking of design choices is the more durable product.","major_comments":[{"comment":"Abstract, Introduction, and Results repeatedly describe the returned front as “globally optimal,” yet the text after Eq. 32 and Alg. 1 explicitly states that forbidding duplicate observations breaks the Markov property and “introduces optimality losses.” For the 6-month Case-b runs (no δix=0), revisits are possible and the claim is therefore overstated. Either (i) restrict the global-optimality language to the frozen-orbit (Case a) 6-month instances where duplicates cannot occur, or (ii) restore Markovian structure (e.g., state augmentation with a visited-set hash or a post-processing exact check on the reported fronts) and quantify the gap. As written, the strongest claim is not fully supported by the algorithm that is run.","section":null},{"comment":"Eq. 21 and the paragraph that follows: the station-keeping model is an empirical linear fit with three coefficients (0.5, 0.5, 1.5 mm/s per m per orbit) calibrated only at a0=6878 km and scaled by a−7/2. Absolute ΔV numbers on the Pareto fronts (Figs. 7–10) and the “meters to tens of m/s” conclusion rest on these free parameters. A short sensitivity study (e.g., ±50% on the three coefficients, or a comparison against a few high-fidelity J2+drag closed-loop runs) is needed so that readers can separate comparative orbit-family conclusions from absolute propellant claims. Without it, the absolute science-per-ΔV numbers remain under-constrained.","section":null},{"comment":"Eqs. 24–25 and Appendix “Science Yield Computations”: the scheduler optimizes a simplified single-Bracewell SNR and Monte-Carlo completeness that omit chopping, multi-baseline LIFE-style architectures, and realistic exozodi/stellar-leakage covariances. That is acceptable for ranking, but the paper’s headline “~10 expected detections” is then an upper-bound ranking metric, not a mission yield forecast. The Results and Conclusion should state this limitation explicitly and avoid equating the optimized objective with expected LIFE/STARI science return.","section":null}],"minor_comments":[{"comment":"Notation: “œ” for orbital elements is nonstandard and hard to read in some fonts; consider the more common “α” or “oe”.","section":null},{"comment":"Eq. (1) and surrounding text: “quasi non-singular relative orbital elements” — a one-line pointer to the exact D’Amico / Koenig convention used would help readers match signs of δλ and δi.","section":null},{"comment":"Fig. 7 caption and sky maps: color scale for fvis is useful; adding the Sun-exclusion boundary as a dashed curve on the α–δ plane would make the winter low-declination gap easier to interpret.","section":null},{"comment":"Table 2 footnote [b]: η_G and η_M citations are given, but η_F and η_K appear without primary references; please supply them.","section":null},{"comment":"Typos / style: “sammples” (Eq. 25 paragraph); “protoplanet” in the Conclusion should be “exoplanet”; “SP ACE” and “OBSERV ABILITY” in headings look like PDF hyphenation artifacts.","section":null},{"comment":"Related work: a brief contrast with classical telescope scheduling (e.g., orienteering / greedy / MILP formulations used for HST/JWST) would better situate the multi-objective DP contribution.","section":null}],"recommendation":"minor_revision","confidential_remarks":"Solid engineering paper appropriate for an AAS/astro-ph.EP venue. The optimality-language overclaim and the untested SK coefficients are the only items that need author attention; neither requires a redesign of the method. I would not block on the simplified SNR model if the authors clearly label it as a ranking metric. Fit to the journal is good."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This is a careful engineering paper that does something useful for the small community designing LEO nulling-interferometer pathfinders. What is actually new is the concrete multi-objective DP that folds linear-formation ROE transfers (with standby relaxation), visibility fractions (Sun exclusion, Earth/solar occultation, baseline oscillation, optional frozen δix), operational dwell, and yield-weighted rewards into a single Pareto policy, plus the numerical fronts for six Sun-synchronous cases. The ROE geometry, linear-formation zero-OPD orbits, and basic ConOps are taken from Hansen and their own prior STARI work; the DP/orienteering machinery is standard. The value is the integration and the comparative results.\n\nThey do the hard parts cleanly. The Bellman update, reachable-set definition, and Pareto pruning are correctly specialized. Fuel lower bounds reduce sensibly for the large-ΔM transfers of interest, and the station-keeping model is at least transparent (empirical coefficients fixed at a0 = 6878 km). Yield comes from an external HWO catalog plus a published SNR formula with Monte-Carlo completeness; that is not circular. The orbit-family comparison (terminator vs noon-midnight, altitude, frozen vs free) is the most useful output: higher-altitude terminator orbits look better on both axes under their models, and the policy is flexible for mid-mission revisits.\n\nSoft spots are real but proportionate. Absolute ΔV and yield numbers rest on those fixed station-keeping coefficients and a simplified SNR; free parameters (dwell window, η rates, B0 clip, SNR = 7, teff,max = 1 day) are listed and not re-tuned to the claimed science return. More importantly, the authors themselves note that forbidding duplicate observations breaks the Markov property, so the returned front is not strictly globally optimal once revisits are properly handled. For a 6-month run that is a minor caveat, but the abstract and strongest claim still say “globally optimal.” No code or data release. None of this sinks the comparative conclusions or the engineering utility.\n\nThis is for mission designers and formation-flying people who need a realistic scheduler, not for pure exoplanet yield theorists. The math and citation pattern look solid; the paper is well above desk-reject. I would send it to referees and would cite the orbit-family comparison and the DP formulation if I were working on STARI-class concepts.","headline":"Solid engineering integration of ROE linear formations, ConOps, and multi-objective DP into a usable Pareto scheduler for LEO nulling pathfinders; absolute yields rest on empirical fuel coeffs and the duplicate-removal step softens the global-optimality claim.","tokens_in":19317,"tokens_out":598,"would_cite":true,"duration_ms":6154,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"A multi-objective dynamic-programming schedule balances fuel and exoplanet yield for LEO nulling interferometers, producing a global Pareto policy under real ConOps constraints.","keywords":["nulling interferometry","formation flying","observation scheduling","multi-objective dynamic programming","relative orbital elements","LEO space telescope","exoplanet yield","Pareto front"],"falsifier":"Fly a high-fidelity closed-loop simulation (or an on-orbit technology demonstrator) of the planned science-standby-transfer sequence and measure whether actual ΔV and achieved optical-path-difference stay within the predicted budgets used to generate the Pareto front.","tokens_in":19334,"feed_emoji":"🛰️","tokens_out":816,"duration_ms":8945,"temperature":0.7,"pith_summary":"Low-Earth-orbit formation-flying interferometers can hunt for biosignatures, but every observation burns fuel and only certain stars are visible at certain times. This paper shows how to turn that trade-off into a solvable multi-objective dynamic program. By folding concept-of-operations constraints, relative-orbit transfer costs, station-keeping burn rates, and visibility windows into a single Pareto-front value function, the method returns the complete set of globally optimal observation sequences from any starting star and remaining propellant budget. Numerical runs on a realistic stellar catalog indicate that a six-month LEO mission can expect up to roughly ten exoplanet detections while spending only a few to a few tens of meters per second of ΔV. The resulting policy is flexible: if a target is re-observed or fuel is diverted, the next best path is already known without re-solving the whole problem.","feed_headline":"LEO interferometers can snag ~10 exoplanets on tens of m/s ΔV","feed_subtitle":"A global Pareto policy balances fuel against scientific yield under real formation-flying constraints.","key_machinery":"The multi-objective Bellman update that builds a Pareto front of cumulative scientific yield versus total ΔV (Eq. 32), initialized from single final observations and pruned of dominated or duplicate paths.","core_discovery":"A multi-objective dynamic-programming scheme that incorporates nominal concept-of-operations constraints, analytical fuel costs for science-to-standby transfers, station-keeping expenditure, and multi-faceted observability windows produces a globally optimal Pareto-front policy for scheduling exoplanetary-system observations with LEO linear-formation interferometers.","pith_inferences":["The same Pareto-DP skeleton can be re-parameterized for eccentric or Sun-Earth L2 formations once their relative-orbit and visibility models are substituted.","Because the policy is already global, ground operators could treat remaining ΔV as a live state variable and replan only when the Markov assumption is broken by a true anomaly change.","Extending the reward from simple yield to a multi-metric science score (biosignature completeness, spectral type balance) would require only a change of the scalar yield term inside the same Bellman update."],"forward_implications":["Mission designers can compare absolute orbits (altitude, LTAN, terminator vs noon-midnight) by reading off their Pareto fronts rather than by single-point analysis.","The same value function immediately supplies the optimal continuation after a revisit or an unplanned fuel expenditure, without re-optimization.","Enforcing the frozen-relative-orbit constraint reduces state dimension and fuel but also shrinks reachable yield; the fronts quantify that trade.","LEO nulling concepts with only meters-to-tens-of-m/s total ΔV budgets remain scientifically competitive for multi-month campaigns."],"fun_headline_variants":["Multi-objective DP yields global Pareto policy for LEO interferometer scheduling","Optimal fuel-science tradeoffs via DP for formation-flying space interferometers","Pareto-front policies schedule exoplanet observations under real LEO constraints","Dynamic programming balances ΔV costs and yield for distributed interferometry","Globally optimal schedules for LEO nulling interferometers from multi-objective DP"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The analytical transfer bound and the empirically calibrated station-keeping formula must accurately predict real propellant use for the linear-formation science and standby orbits under orbital perturbations.","fun_headline_variants_meta":{"raw":{"variants":["Multi-objective DP yields global Pareto policy for LEO interferometer scheduling","Optimal fuel-science tradeoffs via DP for formation-flying space interferometers","Pareto-front policies schedule exoplanet observations under real LEO constraints","Dynamic programming balances ΔV costs and yield for distributed interferometry","Globally optimal schedules for LEO nulling interferometers from multi-objective DP"]},"model":"grok-4.5","effort":"low","cost_usd":0.00613,"raw_usage":{"total_tokens":1525,"prompt_tokens":659,"num_sources_used":0,"completion_tokens":99,"cost_in_usd_ticks":61300000,"prompt_tokens_details":{"text_tokens":659,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":767,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":659,"tokens_out":99,"duration_ms":7603,"temperature":1.0,"reasoning_tokens":767,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-14T08:31:03.049063+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Fly a high-fidelity closed-loop simulation (or an on-orbit technology demonstrator) of the planned science-standby-transfer sequence and measure whether actual ΔV and achieved optical-path-difference stay within the predicted budgets used to generate the Pareto front.","supporting_citations":[],"review_version":1}