NATRO
INITIALIZING PUBLIC RESEARCH INTERFACEVERIFYING DATASET REGISTRYLOADING PROVENANCE GRAPH
BY OL-S-CLOUD
Can a spacecraft health system distinguish physical degradation from degraded observation evidence—and remain calibrated when telemetry quality deteriorates?
yₜ = h(xₜ) + εₜ
xₜ physical state · yₜ observed telemetry · h(·) measurement process · εₜ observation error
εₜ ~ p(ε | rₜ)
rₜ observation reliability conditions the expected observation error
p(xₜ | y₁:ₜ, r₁:ₜ) ∝ p(yₜ | xₜ, rₜ) · p(xₜ | y₁:ₜ₋₁, r₁:ₜ₋₁)
health inference conditions on both telemetry and reliability context
Δcal = ECEbase − ECErel
positive Δcal indicates improved calibration under degraded telemetry
Δrobust = Mrel,degraded − Mbase,degraded
positive Δrobust indicates more graceful degradation under unreliable observations
Pre-results test view: compare calibration error as observation quality worsens.
Pre-results test view: compare performance decay under increasing telemetry degradation.
Fields, units, time axes, quality flags and engineering context.
Reproducible temporal and fault/control cohorts.
Established anomaly baselines before reliability conditioning.
Calibration and robustness under degraded observations.
Canonical historical comparison baseline. Useful for reproducibility, not sufficient alone for the current NATRO scientific claim.
timestampOrdered observation position used to align telemetry and anomaly intervals.
channel_idTelemetry stream identifier for SMAP/MSL benchmark channels.
signal_valueObserved telemetry value at each timestep.
anomaly_sequencePublished anomaly interval used for reproducibility evaluation.
spacecraftMission grouping used to separate SMAP and MSL benchmark sources.
NATRO TERMINAL v0.2 — OL-S-CLOUD
Type help for available commands.
ADAPT · ICESat-2 · Mars 2020 MEDA · SMAP/MSL lineage
OFFICIAL SOURCE →NASA AMESADAPT controlled aerospace fault experiments
OFFICIAL SOURCE →NSIDC DAACICESat-2 ATL02 Version 7 archive and distribution
OFFICIAL SOURCE →NASA PDSMars 2020 MEDA planetary data archive
OFFICIAL SOURCE →ESAOperational spacecraft anomaly benchmark lineage
OFFICIAL SOURCE →Is the physical system behaving abnormally?
Is the observed signal unusual?
Can model confidence itself be trusted?
Can physical degradation be separated from degraded evidence?
Formal benchmarking of diagnostic technologies on the ADAPT electrical-power testbed.
OPEN SOURCE ↗NASA / BENCHMARKEvaluation of ten diagnostic algorithms using ADAPT faults, datasets and common performance metrics.
OPEN SOURCE ↗NASA / PROBABILISTIC DIAGNOSISProbabilistic graphical models and Bayesian-network reasoning for abrupt fault diagnosis.
OPEN SOURCE ↗ESA / SPACECRAFT TELEMETRYLarge-scale, real-life annotated satellite telemetry benchmark developed with ESA spacecraft operations engineers.
OPEN SOURCE ↗ML / CALIBRATIONResearch and engineering contributions across spacecraft telemetry, aerospace and energy systems, machine learning, physical intelligence, explainability and large-scale data infrastructure.
Controlled aerospace testbed rather than operational spacecraft telemetry; valuable because physical faults can be introduced deliberately.
sensor_channelPhysical measurement channel from the aerospace electrical-power testbed.
sensor_valueObserved electrical or component-state measurement.
timeExperiment-relative observation time.
fault_conditionControlled physical fault or altered operating condition.
experiment_runRun identifier supporting controlled system-vs-observation experiments.
Official Version 7 has a published data dictionary. Earthdata authentication is required for file access; metadata discovery is public.
time_ordered_telemetryConverted ATLAS engineering telemetry used in system-level quality control.
instrument_stateInstrument configuration and state information represented in ATL02 telemetry groups.
sensor_characteristicsSensor and instrument characteristics exposed by the Level-1B product.
timingMission and instrument timing needed to align telemetry observations.
calibration_qualityCalibration and product metadata required for defensible reliability analysis.
MEDA separates calibrated and derived products by sensor subsystem and carries native timing and engineering context useful for observation-reliability experiments.
SCLK / LMST / LTSTSpacecraft clock, Local Mean Solar Time and Local True Solar Time references.
ATSCalibrated Air Temperature Sensor observations.
PSAtmospheric pressure sensor products.
RHSRelative Humidity Sensor observations and derived context.
TIRSThermal Infrared Sensor measurements including surface/thermal context.
WINDWind speed and direction products.
RDSRadiation and Dust Sensor measurements.
ENGEngineering telemetry associated with MEDA instrument operation.
Independent external validation source with real operational telemetry, anomaly labels, rare-event context and telecommands.
telemetry_channelOperational telemetry stream belonging to a spacecraft subsystem.
timestampTime reference used to align telemetry, rare events and anomaly intervals.
subsystemSubsystem grouping for cross-channel reasoning.
anomaly_eventPublished anomaly annotations for benchmark evaluation.
rare_eventRare but potentially non-fault operational events.
telecommand_contextCommand activity that can help distinguish operational changes from faults.
Foundational modern reference for confidence calibration and expected calibration error.