juliensimon/constellation-tle-latest

Constellation TLEs — 18 Satellite Constellations

Credit: NASA

Part of the Orbital Mechanics Datasets collection on Hugging Face.

Daily Two-Line Element (TLE) snapshots for 17 satellite constellations totaling 1,747 satellites, sourced from CelesTrak. Each constellation is available as a separate config for independent loading.

Constellations

Constellation	Operator	Orbit	Satellites
OneWeb	Eutelsat OneWeb	LEO	651
Kuiper	Amazon	LEO	210
Qianfan (G60)	Shanghai Spacecom	LEO	108
Hulianwang (GuoWang)	China SatNet	LEO	159
Iridium NEXT	Iridium	LEO	80
Globalstar	Globalstar	LEO	28
ORBCOMM	ORBCOMM	LEO	15
Planet Labs	Planet Labs	LEO	137
Spire Global	Spire Global	LEO	77
GPS (NAVSTAR)	USSF	MEO	39
Galileo	EU/ESA	MEO	50
BeiDou	CNSA	MEO/GEO	56
SBAS	Various	GEO	6
SES	SES	GEO	70
Intelsat	Intelsat	GEO	27
Eutelsat	Eutelsat	GEO	29
Telesat	Telesat	GEO/LEO	5

Total: 1,747 satellites — 1,465 LEO, 145 MEO, 137 GEO

Dataset description

Two-Line Element sets (TLEs) are the standard format for representing satellite orbital elements, developed by NORAD and used universally with the SGP4/SDP4 propagation model. This dataset provides daily-fresh TLEs for every major non-Starlink constellation in orbit, organized by operator for easy access.

The collection spans four categories:

GNSS navigation (GPS, Galileo, BeiDou, GLONASS, SBAS) — the backbone of global positioning in MEO/GEO, critical for timing applications, geodesy, and as reference orbits for propagation model validation.

LEO broadband (OneWeb, Kuiper, Qianfan, Hulianwang) — the next wave of mega-constellations competing with Starlink. OneWeb is fully deployed at 1,200 km; Amazon's Kuiper and China's Qianfan/Hulianwang are in early deployment phases.

LEO communications and IoT (Iridium NEXT, Globalstar, ORBCOMM) — established mobile satellite service providers operating in low Earth orbit.

Earth observation (Planet Labs, Spire Global) — the largest commercial imaging and weather monitoring fleets, operating in sun-synchronous orbits.

GEO communications (SES, Intelsat, Eutelsat, Telesat) — legacy geostationary operators providing broadcast, broadband, and government services from 35,786 km.

Raw .tle files are provided alongside Parquet for maximum compatibility: orbit propagation libraries like python-sgp4, orekit, and STK consume the standard three-line TLE format directly. Because TLE accuracy degrades rapidly (especially for LEO objects), daily updates are essential for operational applications.

Starlink TLEs are published separately in starlink-tle-latest due to the constellation's size (7,000+ satellites).

Raw TLE files

For applications that consume standard 3-line TLE format (e.g., SGP4 propagators):

• data/oneweb.tle — OneWeb
• data/kuiper.tle — Kuiper
• data/qianfan.tle — Qianfan (G60)
• data/hulianwang.tle — Hulianwang (GuoWang)
• data/iridium.tle — Iridium NEXT
• data/globalstar.tle — Globalstar
• data/orbcomm.tle — ORBCOMM
• data/planet.tle — Planet Labs
• data/spire.tle — Spire Global
• data/gps.tle — GPS (NAVSTAR)
• data/galileo.tle — Galileo
• data/beidou.tle — BeiDou
• data/sbas.tle — SBAS
• data/ses.tle — SES
• data/intelsat.tle — Intelsat
• data/eutelsat.tle — Eutelsat
• data/telesat.tle — Telesat

Schema (all configs)

Column	Type	Description
name	string	Satellite name (e.g., "ONEWEB-0012", "NAVSTAR 78", "GALILEO 27")
line1	string	TLE line 1: epoch, drag term, BSTAR, element set number
line2	string	TLE line 2: inclination, RAAN, eccentricity, arg of perigee, mean anomaly, mean motion

Quick stats

• 1,747 satellites across 17 constellations
• 1,465 LEO + 145 MEO + 137 GEO
• Snapshot: 2026-04-06 14:00 UTC

Usage

from datasets import load_dataset

# Load a specific constellation
gps = load_dataset("juliensimon/constellation-tle-latest", "gps", split="train")
oneweb = load_dataset("juliensimon/constellation-tle-latest", "oneweb", split="train")
galileo = load_dataset("juliensimon/constellation-tle-latest", "galileo", split="train")

# Use with sgp4 library for orbit propagation
from sgp4.api import Satrec
sat = Satrec.twoline2rv(gps[0]["line1"], gps[0]["line2"])

# Load all GNSS constellations
import pandas as pd
gnss = pd.concat([
    load_dataset("juliensimon/constellation-tle-latest", c, split="train").to_pandas()
    for c in ["gps", "galileo", "beidou", "glonass", "sbas"]
])
print(f"{len(gnss)} GNSS satellites")

# Compare constellation sizes
for config in ["oneweb", "iridium", "planet", "spire"]:
    ds = load_dataset("juliensimon/constellation-tle-latest", config, split="train")
    print(f"{config}: {len(ds)} satellites")

Data source

CelesTrak (Dr. T.S. Kelso), mirroring NORAD/18th Space Defense Squadron data. No authentication required.

Update schedule

Daily at 05:30 UTC via GitHub Actions.

Related datasets

• starlink-tle-latest — Daily Starlink + GPS TLEs (dedicated dataset for the largest constellation)
• space-track-tle-history — 238M+ historical TLEs for every cataloged object (1959–present)
• constellation-census — Parsed orbital elements with status classification for all constellations
• space-track-satcat — Complete NORAD satellite catalog (68K+ objects)
• ucs-satellite-database — Active satellites with purpose, operator, and orbit metadata

Pipeline

Source code: juliensimon/space-datasets

Support

If you find this dataset useful, please give it a ❤️ on the dataset page and share feedback in the Community tab! Also consider giving a ⭐️ to the space-datasets repo.

Citation

@dataset{constellation_tle_latest,
  author = {Simon, Julien},
  title = {Constellation TLEs — 18 Satellite Constellations},
  year = {2026},
  publisher = {Hugging Face},
  url = {https://huggingface.co/datasets/juliensimon/constellation-tle-latest},
  note = {Based on NORAD data via CelesTrak (Dr. T.S. Kelso)}
}

License

CC-BY-4.0