{"claim": "Mobility productivity rose steeply across the Mars rover fleet (MER, Curiosity, Perseverance) and the autonomy and hardware channels are described together and never separated, so the problem the dissertation isolates is real: a documented fleet-wide productivity rise with two confounded candidate drivers.", "evidence": [{"source": "Crisp, Adler, Matijevic, Squyres, Arvidson, Kass, 'Mars Exploration Rover mission,' J. Geophysical Research: Planets (2003)", "doi_or_url": "https://doi.org/10.1029/2002je002038", "grade": "A"}, {"source": "Grotzinger et al., 'Mars Science Laboratory Mission and Science Investigation,' Space Science Reviews (2012)", "doi_or_url": "https://doi.org/10.1007/s11214-012-9892-2", "grade": "A"}, {"source": "Farley et al., 'Overview and Results From the Mars 2020 Perseverance Rover First Science Campaign on the Jezero Crater Floor,' J. Geophysical Research: Planets (2023)", "doi_or_url": "https://doi.org/10.1029/2022je007613", "grade": "A"}], "facet": "empirics", "chapter": "ch3_literature_review", "subclaim": "real"}
{"claim": "Vasavada et al. document the Mars Science Laboratory mobility record from Bradbury Landing onward, supplying the mid-generation (Curiosity) leg of the cross-fleet productivity description that establishes the rise is observed across all three generations rather than at a single rover.", "evidence": [{"source": "Vasavada et al., 'Overview of the Mars Science Laboratory mission: Bradbury Landing to Yellowknife Bay and beyond,' J. Geophysical Research: Planets (2014)", "doi_or_url": "https://doi.org/10.1002/2014je004622", "grade": "A"}], "facet": "empirics", "chapter": "ch3_literature_review", "subclaim": "real"}
{"claim": "Ground-in-the-loop operation is the binding scarcity in space-robot mission practice and onboard autonomy is the lever that relaxes it; the survey states the payoff is described but never causally identified, which is precisely the gap the dissertation addresses and which makes the productivity question decision-relevant (material).", "evidence": [{"source": "Gao and Chien, 'Autonomy for Space Robots: Past, Present, and Future,' Current Robotics Reports (2021)", "doi_or_url": "https://doi.org/10.1007/s43154-021-00057-2", "grade": "A"}], "facet": "mechanism", "chapter": "ch3_literature_review", "subclaim": "material"}
{"claim": "Sols are the binding scarcity for Mars Sample Return caching, instrument placement, and the daily planning cycle, so a per-sol mobility-productivity gain is decision-relevant; the Perseverance first-science-campaign record shows the caching-and-sampling campaign whose throughput the sol budget governs, making the channel question material to mission design.", "evidence": [{"source": "Farley et al., 'Overview and Results From the Mars 2020 Perseverance Rover First Science Campaign on the Jezero Crater Floor,' J. Geophysical Research: Planets (2023)", "doi_or_url": "https://doi.org/10.1029/2022je007613", "grade": "A"}, {"source": "Golombek et al., 'Terrain physical properties derived from orbital data and the first 360 sols of MSL Curiosity observations in Gale Crater,' J. Geophysical Research: Planets (2014)", "doi_or_url": "https://doi.org/10.1002/2013je004605", "grade": "A"}], "facet": "economics", "chapter": "ch3_literature_review", "subclaim": "material"}
{"claim": "Terrain physical properties (load-bearing strength, slip, sinkage) are derivable a priori from orbital data and early-sol observations, which supplies the constructed terrain-class crosswalk the design conditions on and grounds the claim that an a-priori terrain block can be defined before the drive plan.", "evidence": [{"source": "Golombek et al., 'Terrain physical properties derived from orbital data and the first 360 sols of MSL Curiosity observations in Gale Crater,' J. Geophysical Research: Planets (2014)", "doi_or_url": "https://doi.org/10.1002/2013je004605", "grade": "A"}], "facet": "measurement", "chapter": "ch4_data_and_measurement", "subclaim": "real"}
{"claim": "Credible causal inference begins by naming the source of exogenous variation; research design, not functional-form adjustment, removes bias, and selection-on-observables rests on the conditional-independence assumption remaining defensible, so conditioning that absorbs the treatment-relevant terrain variation undermines identification. This is the standard the within-rover autonomous-fraction first stage must meet.", "evidence": [{"source": "Angrist & Pischke, 'The Credibility Revolution in Empirical Economics,' J. Economic Perspectives 24(2) (2010)", "doi_or_url": "https://doi.org/10.1257/jep.24.2.3", "grade": "A"}, {"source": "Angrist & Pischke, Mostly Harmless Econometrics: An Empiricist's Companion (2009)", "doi_or_url": "https://doi.org/10.1515/9781400829828", "grade": "A"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "Bad controls (variables themselves caused by, or forecasting, treatment or outcome) reintroduce bias; valid controls must be pre-determined relative to treatment, and coefficient stability as the suspect block is added versus omitted is the right diagnostic. This disciplines the requirement that the terrain covariate be timestamped before the drive plan rather than an outcome forecast.", "evidence": [{"source": "Angrist & Pischke, Mostly Harmless Econometrics: An Empiricist's Companion (2009), bad-controls diagnostic", "doi_or_url": "https://doi.org/10.1515/9781400829828", "grade": "A"}], "facet": "measurement", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "Asymptotic cluster-robust inference is unreliable with a small number of clusters, and the wild bootstrap for few (treated) clusters is a distinct, size-distorted regime requiring specialized procedures; with three between-rover clusters the default wild-cluster-bootstrap p-values are not credible, so a design-specific Monte Carlo of the true rejection rate is the appropriate demand before the between-rover coefficient is read as a hypothesis test.", "evidence": [{"source": "Cameron, Gelbach & Miller, 'Bootstrap-Based Improvements for Inference with Clustered Errors'", "doi_or_url": "https://doi.org/10.2139/ssrn.956890", "grade": "A"}, {"source": "MacKinnon & Webb, 'The wild bootstrap for few (treated) clusters,' The Econometrics Journal 20(2) (2017)", "doi_or_url": "https://doi.org/10.1111/ectj.12107", "grade": "A"}], "facet": "empirics", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "Selection-on-observables identifies a causal effect only when conditional independence holds given the conditioning set; omitting a covariate (such as a drive-sol time-pressure selector) that drives both the autonomous-fraction choice and meters-per-sol reintroduces omitted-variable bias, so conditioning on terrain alone does not deliver the CIA the within-rover design needs.", "evidence": [{"source": "Angrist & Pischke, Mostly Harmless Econometrics (2009); Mastering 'Metrics: The Path from Cause to Effect (2014)", "doi_or_url": "https://doi.org/10.1515/9781400829828", "grade": "A"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "Credible causal inference rests on an identified contrast (randomized or quasi-experimental variation), not on covariate-adjusted explained variance; under near-perfect collinearity between the autonomy and hardware blocks the between-block partition is order-dependent and not identified, so a horse-race verdict resting on incremental R-squared is an artifact of entry order rather than an identified estimand.", "evidence": [{"source": "Angrist & Pischke, Mostly Harmless Econometrics: An Empiricist's Companion (2009)", "doi_or_url": "https://doi.org/10.1515/9781400829828", "grade": "A"}], "facet": "measurement", "chapter": "ch5_research_design", "subclaim": "alternatives"}
{"claim": "Design-based inference requires falsification checks (including placebo and zero-by-construction negative controls) that can in principle reject the hypothesized channel; the candidate's existing falsification checks are insufficient because they are all internal to H1 rather than negative controls on a margin where the software channel provably cannot operate.", "evidence": [{"source": "Angrist & Pischke, Mostly Harmless Econometrics (2009), regression and bad controls; Mastering 'Metrics (2014)", "doi_or_url": "https://doi.org/10.1515/9781400829828", "grade": "A"}], "facet": "empirics", "chapter": "ch6_analysis_plan", "subclaim": "residual_risk"}
{"claim": "For a continuous treatment, treatment-on-the-treated-type parameters are identified under an analogous parallel-trends assumption, but comparing parameters across dose levels does not follow from parallel trends because units selecting into higher doses can differ systematically; reading the within-rover autonomous-fraction slope as marginal meters-per-sol of more autonomy is therefore a forbidden cross-dose comparison unless a stronger no-selection-into-dose assumption is stated or H1 is restricted to TOT at observed doses.", "evidence": [{"source": "Callaway, Goodman-Bacon & Sant'Anna, 'Difference-in-Differences with a Continuous Treatment' (2024 working paper)", "doi_or_url": "https://doi.org/10.2139/ssrn.4716682", "grade": "B"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "A doubly-robust difference-in-differences estimator for the ATT is consistent if either the propensity-score (dose-assignment) model or the outcome-regression model is correctly specified; when the ground team chooses a higher autonomous fraction on terrain it judges suitable, identification needs covariate-conditional parallel trends and double robustness supplies a defensible consistency route at observed doses.", "evidence": [{"source": "Sant'Anna & Zhao, 'Doubly robust difference-in-differences estimators,' Journal of Econometrics (2020)", "doi_or_url": "https://doi.org/10.1016/j.jeconom.2020.06.003", "grade": "A"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "alternatives"}
{"claim": "The two-way fixed-effects estimator equals a weighted average of all two-group two-period difference-in-differences comparisons; under heterogeneous effects with staggered timing, comparisons using already-treated units as controls can carry negative weights and the pooled coefficient can be sign-flipped relative to every underlying effect, which is exactly the regime created by terrain-heterogeneous, within-mission-growing, staggered generation switch-on.", "evidence": [{"source": "Goodman-Bacon, 'Difference-in-differences with variation in treatment timing,' Journal of Econometrics (2021)", "doi_or_url": "https://doi.org/10.1016/j.jeconom.2021.03.014", "grade": "A"}], "facet": "empirics", "chapter": "ch5_research_design", "subclaim": "alternatives"}
{"claim": "The disciplined replacement for contaminated TWFE is disaggregated ATT(g,t) building blocks identified off never-treated or not-yet-treated controls (already-treated units excluded from the control pool), aggregated transparently with researcher-chosen weights; with no never-treated rover, each generation's effect must be identified off not-yet-treated segments so no later-generation segment enters another generation's control comparison.", "evidence": [{"source": "Callaway & Sant'Anna, 'Difference-in-Differences with multiple time periods,' Journal of Econometrics (2021)", "doi_or_url": "https://doi.org/10.1016/j.jeconom.2020.12.001", "grade": "A"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "alternatives"}
{"claim": "A flat pre-trend is necessary but not sufficient: pre-trends tests have low power and conditioning the analysis on having passed one can itself distort estimates and worsen bias, so a single flat-pretrend plot is inadmissible as the parallel-trends defense and a sensitivity analysis bounding the estimate under plausible violations must be reported instead.", "evidence": [{"source": "Roth, 'Pretest with Caution: Event-Study Estimates after Testing for Parallel Trends,' American Economic Review: Insights (2022)", "doi_or_url": "https://doi.org/10.1257/aeri.20210236", "grade": "A"}, {"source": "Roth, Sant'Anna, Bilinski & Poe, 'What's trending in difference-in-differences? A synthesis of the recent econometrics literature,' Journal of Econometrics (2023)", "doi_or_url": "https://doi.org/10.1016/j.jeconom.2023.03.008", "grade": "A"}], "facet": "identification", "chapter": "ch6_analysis_plan", "subclaim": "residual_risk"}
{"claim": "Callaway & Sant'Anna identification rests jointly on parallel trends and a no-anticipation assumption, with pre-treatment ATT(g,t) leads as the prescribed placebo diagnostic; a trust-ramped autonomous fraction driven by realized success makes the dose a function of the outcome path, an anticipation/feedback violation distinct from the bad-controls rule, testable by regressing the fraction on lagged meters-per-sol with a flat pre-dose placebo as the clearing check.", "evidence": [{"source": "Callaway & Sant'Anna, 'Difference-in-Differences with multiple time periods,' Journal of Econometrics (2021)", "doi_or_url": "https://doi.org/10.1016/j.jeconom.2020.12.001", "grade": "A"}, {"source": "Roth, 'Pretest with Caution: Event-Study Estimates after Testing for Parallel Trends,' AER: Insights (2022)", "doi_or_url": "https://doi.org/10.1257/aeri.20210236", "grade": "A"}], "facet": "empirics", "chapter": "ch6_analysis_plan", "subclaim": "residual_risk"}
{"claim": "The institutional cost of certifying an in-flight rover autonomy upgrade is real and measurable from records already in scope: the MER mobility flight-software surface-test/validation regime, the measured directed-versus-autonomous throughput tradeoff, the documented global-planner software-integration-and-surface-test burden, and the certified in-flight Enhanced AutoNav upload on Perseverance together supply a certification-cost term, so the software channel is not free and the cost must enter the software-versus-mass ledger.", "evidence": [{"source": "Biesiadecki & Maimone, 'The Mars Exploration Rover Surface Mobility Flight Software: Driving Ambition,' IEEE Aerospace Conf. 2006", "doi_or_url": "https://doi.org/10.1109/aero.2006.1655723", "grade": "A"}, {"source": "Biesiadecki et al., 'Tradeoffs Between Directed and Autonomous Driving on the Mars Exploration Rovers,' Int. J. Robotics Research (2007)", "doi_or_url": "https://doi.org/10.1177/0278364907073777", "grade": "A"}, {"source": "'Global planning on the Mars Exploration Rovers: Software integration and surface test,' J. Field Robotics (2009)", "doi_or_url": "https://doi.org/10.1002/rob.20287", "grade": "A"}, {"source": "Verma et al., 'Enhanced Autonomous Navigation on the Perseverance Mars Rover,' IEEE Trans. Field Robotics (2025)", "doi_or_url": "https://doi.org/10.1109/tfr.2025.3636366", "grade": "A"}], "facet": "economics", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "A mission-independent capability axis on which a generation discontinuity could be defined exists and is measured per-sol across the fleet: drive distance per sol and the directed-versus-autonomous drive fraction are reported for MER, tracked longitudinally for Curiosity, and stepped by Enhanced AutoNav on Perseverance; the measurement substrate is grounded even though no retrieved source defines the G1/G2/G3 bright line as a measured discontinuity on that axis.", "evidence": [{"source": "Biesiadecki et al., 'Tradeoffs Between Directed and Autonomous Driving on the Mars Exploration Rovers,' Int. J. Robotics Research (2007)", "doi_or_url": "https://doi.org/10.1177/0278364907073777", "grade": "A"}, {"source": "Rankin et al., 'Driving Curiosity: Mars Rover Mobility Trends During the First Seven Years,' IEEE Aerospace Conf. 2020", "doi_or_url": "https://doi.org/10.1109/aero47225.2020.9172469", "grade": "A"}, {"source": "Verma et al., 'Enhanced Autonomous Navigation on the Perseverance Mars Rover,' IEEE Trans. Field Robotics (2025)", "doi_or_url": "https://doi.org/10.1109/tfr.2025.3636366", "grade": "A"}], "facet": "measurement", "chapter": "ch4_data_and_measurement", "subclaim": "residual_risk"}
{"claim": "Per-sol rover productivity is structurally gated by the uplink/downlink cycle (each sol's downlink determines the next sol's uplink command set) and autonomous-planner operations concepts were built specifically because the ground-in-the-loop planning cycle bounds productivity, so pooling a Perseverance autonomous-drive sol with an Opportunity blind-commanded sol without an operations-cadence covariate risks attributing to autonomy a per-sol difference that is partly a cadence artifact; the directed-versus-autonomous distinction is measured per sol so the two signals can be disentangled.", "evidence": [{"source": "'Relating downlink products to uplink commands in Mars rover operations,' IEEE Aerospace Conf. 2002", "doi_or_url": "https://doi.org/10.1109/aero.2002.1035301", "grade": "A"}, {"source": "'Innovative Rover Operations Concepts - Autonomous Planner (IRONCAP),' AIAA SpaceOps 2012", "doi_or_url": "https://doi.org/10.2514/6.2012-1294460", "grade": "B"}, {"source": "Biesiadecki et al., 'Tradeoffs Between Directed and Autonomous Driving on the Mars Exploration Rovers,' Int. J. Robotics Research (2007)", "doi_or_url": "https://doi.org/10.1177/0278364907073777", "grade": "A"}], "facet": "measurement", "chapter": "ch4_data_and_measurement", "subclaim": "residual_risk"}
{"claim": "Off-Earth timekeeping (the Mars sol / Darian system) is designed, site-local reference-frame infrastructure rather than a neutral universal denominator, so a meters-per-sol value silently carries a rover-specific time frame that must be declared and converted before any cross-site pooling.", "evidence": [{"source": "Gangale, 'The Architecture of Time, Part 2: The Darian System for Mars,' SAE Technical Paper 2006-01-2249", "doi_or_url": "https://doi.org/10.4271/2006-01-2249", "grade": "B"}], "facet": "measurement", "chapter": "ch4_data_and_measurement", "subclaim": "residual_risk"}
{"claim": "Lunar rover autonomous path planning is an active research frontier (the candidate's own cited reference), confirming a plausible cross-platform replication target exists, but it is algorithm development rather than a lunar traverse-productivity estimation archive, so it does not supply the off-Mars meters-per-sol dataset a pre-registered cross-platform transfer test would re-estimate on.", "evidence": [{"source": "Chen, Jackson, Allard, Beltrame, 'Path planning algorithm for a South Pole lunar rover mission,' Acta Astronautica 237 (2025)", "doi_or_url": "https://doi.org/10.1016/j.actaastro.2025.07.059", "grade": "A"}], "facet": "identification", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "Perseverance design replicates Curiosity except where new objectives mandated change, and the key upgrades were bundled together (more robust tires, new engineering cameras, a new computer dedicated to image processing, and a more efficient AutoNav), so the autonomy generation rode in with compute and mass on a richer mission; the between-rover G1/G2/G3 contrast is therefore physically confounded with the mission-class budget unless a measured budget control is entered.", "evidence": [{"source": "'Planning for a Martian Road Trip - The Mars 2020 Mobility Systems Design' (NASA NTRS 20230005676)", "doi_or_url": "https://ntrs.nasa.gov/citations/20230005676", "grade": "C"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "A within-mission, post-landing autonomy increment on a fixed platform exists in the NTRS record: directed-versus-autonomous driving shifted as onboard planning capability deepened on the same deployed MER rovers, providing a within-budget contrast that holds mission class, mass, compute, and downlink fixed and is the variation the dissertation makes primary against the funded-mission-class-selection rival.", "evidence": [{"source": "'Tradeoffs between directed and autonomous driving on the Mars Exploration Rover' (NASA NTRS 20060042880)", "doi_or_url": "https://ntrs.nasa.gov/citations/20060042880", "grade": "C"}], "facet": "rival", "chapter": "ch5_research_design", "subclaim": "alternatives"}
{"claim": "Autonomous on-board navigation (visual odometry) was demonstrated and characterized over two years on the Mars Exploration Rovers, establishing that the productivity upside of rover autonomy is measurable from the operational/telemetry record; this anchors the gain side of a net-effect ledger.", "evidence": [{"source": "Maimone, Cheng et al., 'Two years of Visual Odometry on the Mars Exploration Rovers,' Journal of Field Robotics (2007)", "doi_or_url": "https://doi.org/10.1002/rob.20184", "grade": "A"}], "facet": "empirics", "chapter": "ch3_literature_review", "subclaim": "real"}
{"claim": "Verification and validation of autonomous systems for space exploration is a recognized, unsolved hard problem, which is the mechanism by which an over-trusted or under-verified autonomy command set produces aborts, safe-mode entries, and propagated faults; a doctrine estimated on the gross upside alone cannot establish the sign of the net effect because this downside loop is unmeasured.", "evidence": [{"source": "Brat & Jonsson, 'Challenges in verification and validation of autonomous systems for space exploration,' IJCNN (2005)", "doi_or_url": "https://doi.org/10.1109/ijcnn.2005.1556387", "grade": "A"}], "facet": "empirics", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "An independent propositional-base observable exists and is measurable separately from the autonomy-generation coefficient: onboard slip prediction migrated from no model, to image-learned visual slip prediction, to mechanics-based wheel slip/sinkage estimation, a deepening theory of why terrain behaves as it does, so a per-generation model-accuracy series would test Mokyr extensibility rather than letting the coefficient stand in for a widening knowledge base.", "evidence": [{"source": "Angelova, Matthies, Helmick, Perona, 'Slip Prediction Using Visual Information,' Robotics: Science and Systems (2006)", "doi_or_url": "https://doi.org/10.15607/rss.2006.ii.014", "grade": "A"}, {"source": "'Methods for Wheel Slip and Sinkage Estimation in Mobile Robots' (2010)", "doi_or_url": "https://doi.org/10.5772/9279", "grade": "B"}], "facet": "measurement", "chapter": "ch2_theoretical_framework", "subclaim": "residual_risk"}
{"claim": "Adaptive and intelligent (model-driven, extensible) rover navigation is surveyed as a distinct capability axis from raw autonomous-traverse utilization, so a within-rover autonomous-drive-fraction coefficient by itself does not evidence knowledge-base widening and is observationally equivalent to a fixed technique utilized more aggressively (prescriptive habituation).", "evidence": [{"source": "'Adaptive and intelligent navigation of autonomous planetary rovers - A survey' (2017)", "doi_or_url": "https://doi.org/10.1109/ahs.2017.8046384", "grade": "B"}, {"source": "Angelova et al., 'Slip Prediction Using Visual Information,' RSS 2006", "doi_or_url": "https://doi.org/10.15607/rss.2006.ii.014", "grade": "A"}], "facet": "mechanism", "chapter": "ch2_theoretical_framework", "subclaim": "residual_risk"}
{"claim": "Onboard autonomy for space robots advances through discrete, validation-gated software stages on operating assets rather than continuous frictionless upload, supporting the access-cost (stickiness) prediction over the design's frictionless in-principle retrofit claim; the realized access cost is the calendar lag from development to operational-on-surface in each asset's flight-software version history.", "evidence": [{"source": "Gao & Chien, 'Autonomy for Space Robots: Past, Present, and Future,' Current Robotics Reports (2021)", "doi_or_url": "https://doi.org/10.1007/s43154-021-00057-2", "grade": "A"}, {"source": "NASA/DOD Aerospace Knowledge Diffusion Research Project: US STI Policy (NTRS)", "doi_or_url": "https://ntrs.nasa.gov/citations/19960052732", "grade": "B"}], "facet": "economics", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "Major rover autonomy and imaging capability increments are tied to new rover generations delivered at launch (Mars 2020 next-generation engineering-camera/imaging system; the MSL eight-year surface-operations record), so the generational pattern suggests most step changes shipped at launch, which would collapse the software/hardware retrofit asymmetry if the per-asset retrofit fraction is not estimated.", "evidence": [{"source": "'The Mars 2020 Engineering Cameras and Microphone on the Perseverance Rover' (2020)", "doi_or_url": "https://doi.org/10.1007/s11214-020-00765-9", "grade": "A"}], "facet": "economics", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "Surface-autonomy certification is formalizable as a structured verification-and-validation campaign with non-trivial assurance burden, establishing that a retrofit-path certification cost is a legitimate non-zero ledger entry rather than an assumed-near-zero quantity; this grounds the objection that the candidate's software-versus-mass ledger omits an institutional access cost, though the Mars rover-specific magnitude remains unmeasured.", "evidence": [{"source": "Mendoza et al., 'A Method for Lunar Surface Autonomy Certification: Application to a Construction Pathfinder Mission,' Aerospace 12(12):1115 (2025)", "doi_or_url": "https://doi.org/10.3390/aerospace12121115", "grade": "A"}], "facet": "economics", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "Trust in automation should be calibrated to true reliability; miscalibration runs as misuse (over-reliance beyond competence) and disuse (under-reliance), making operator activation of automation a voluntary reliance choice distinct from the automation's competence, so a positive autonomous-fraction coefficient is confounded by the ground team learning to trust an unchanged machine and is not, on its own, evidence of machine hazard competence.", "evidence": [{"source": "Parasuraman & Riley, 'Humans and Automation: Use, Misuse, Disuse, Abuse,' Human Factors 39(2):230-253 (1997)", "doi_or_url": "https://doi.org/10.1518/001872097778543886", "grade": "A"}], "facet": "identification", "chapter": "ch2_theoretical_framework", "subclaim": "residual_risk"}
{"claim": "High automation reliability predictably degrades operator monitoring of the automated function (automation-induced complacency), a robust dosage-dependent attentional-reallocation effect; higher monitor reliability measurably reduced operators' detection of automation failures, supporting the prediction that a trusted AutoNav frees planning-cycle attention and induces longer, bolder commanded drives, inflating meters-per-sol through the commanding channel rather than the hazard-clearing channel.", "evidence": [{"source": "Parasuraman & Manzey, 'Complacency and Bias in Human Use of Automation: An Attentional Integration,' Human Factors 52(3):381-410 (2010)", "doi_or_url": "https://doi.org/10.1177/0018720810376055", "grade": "A"}, {"source": "Parasuraman, Molloy & Singh, 'Performance Consequences of Automation-Induced Complacency,' Int. J. of Aviation Psychology 3(1):1-23 (1993)", "doi_or_url": "https://doi.org/10.1207/s15327108ijap0301_1", "grade": "A"}], "facet": "mechanism", "chapter": "ch2_theoretical_framework", "subclaim": "residual_risk"}
{"claim": "Automation reliance generates errors of omission (failing to act on conditions the automation did not flag) and errors of commission (acting on the automation against contrary evidence), so a meters-per-sol outcome counting only successful traverse omits the commission/fault tail; netting autonomy-attributable fault, safing, and abort sols is the correct robustness test and a trust-calibration account predicts the coefficient would shrink.", "evidence": [{"source": "Parasuraman & Manzey, 'Complacency and Bias in Human Use of Automation: An Attentional Integration,' Human Factors 52(3):381-410 (2010)", "doi_or_url": "https://doi.org/10.1177/0018720810376055", "grade": "A"}], "facet": "measurement", "chapter": "ch2_theoretical_framework", "subclaim": "residual_risk"}
{"claim": "Automation applies to four distinct stages of human information processing (information acquisition, information analysis, decision/action selection, action implementation), each automatable to a different level, so a single continuous autonomy scalar that collapses a stage-4 timing change (think-while-driving) with a stage-2/3 capability change is mis-specified and not interpretable as how much the machine is doing.", "evidence": [{"source": "Parasuraman, Sheridan & Wickens, 'A model for types and levels of human interaction with automation,' IEEE Trans. SMC-A (2000)", "doi_or_url": "https://doi.org/10.1109/3468.844354", "grade": "A"}], "facet": "mechanism", "chapter": "ch2_theoretical_framework", "subclaim": "residual_risk"}
{"claim": "Operator reliance is dynamic, not constant: vigilance and workload are time-varying and a high false-alarm threshold drives operators to disuse the alarm (cry-wolf), so a ground-team conservatism response co-moving with the autonomous fraction is an omitted dynamic confounder that the additive constant-ideal-human specification ignores; the candidate must demonstrate the autonomous-fraction coefficient is separable from this reliance-response channel.", "evidence": [{"source": "Parasuraman, 'Neuroergonomics: Research and practice,' Theoretical Issues in Ergonomics Science (2003)", "doi_or_url": "https://doi.org/10.1080/14639220210199753", "grade": "A"}, {"source": "Parasuraman & Riley, 'Humans and Automation: Use, Misuse, Disuse, Abuse,' Human Factors (1997)", "doi_or_url": "https://doi.org/10.1518/001872097778543886", "grade": "A"}], "facet": "identification", "chapter": "ch5_research_design", "subclaim": "residual_risk"}
{"claim": "System state lives in stocks (accumulations) that change only through flows (rates), so meters-per-sol is a flow while the binding scarcity is a stock (one irreplaceable vehicle plus a finite non-renewable stock of mission-sols inside an MSR launch window); a correct loss function for a fleet-wide doctrine is an integral over the residual horizon with the catastrophic branch driving the integrand to zero at the abort sol, not a point estimate of forgone rate, and stock buffers rank above parameters on the leverage hierarchy.", "evidence": [{"source": "Meadows, 'Leverage Points: Places to Intervene in a System' (1999)", "doi_or_url": "https://doi.org/10.4324/9781849773386-15", "grade": "A"}], "facet": "economics", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "Optimize-for-trusted-autonomy-fleet-wide changes the goal of the operations system, a high-leverage tier that amplifies a directional error, and its mechanism of harm is balancing-loop erosion: ground-commanding skill and the conservative reserve margin are stabilizing stocks a reinforcing reward-higher-autonomy loop can draw down, detectable in telemetry as rising autonomous-fraction co-occurring with falling ground-intervention frequency and thinning commanded margin over sols.", "evidence": [{"source": "Meadows, 'Leverage Points: Places to Intervene in a System' (1999)", "doi_or_url": "https://doi.org/10.4324/9781849773386-15", "grade": "A"}], "facet": "governance", "chapter": "ch7_discussion", "subclaim": "residual_risk"}
{"claim": "Goswami & Garretson hold that capability is endogenous to elite preference and program funding, so an autonomy-generation contrast that cannot be separated from the funded mission class it rode in on is a preference/budget index wearing a capability label; this grounds the demand for a pre-registered observable and falsifier and for a measured budget control before the between-rover contrast is read causally.", "evidence": [{"source": "Goswami & Garretson, Scramble for the Skies (preference-indicator / four-driver decomposition framework)", "doi_or_url": "https://doi.org/10.5040/9781978730243.ch-005", "grade": "A"}], "facet": "rival", "chapter": "ch5_research_design", "subclaim": "alternatives"}