Sodern launches Astradia, a star tracker for GNSS-denied navigation

by sorenjanon 6/5/2025, 11:20 AMwith 47 comments

by perihelionson 6/5/2025, 11:39 AM

I wonder if it'd make sense these days to put optical navigation beacons in space. This OP solution relies on inertial navigation, because, the star-tracking part is only a reference for orientation in space, which is an incomplete coordinate set. But if you were to put *artificial* stars close to Earth, you'd gain an additional parallax measurement, and obtain a complete basis.

by Vox_Leoneon 6/5/2025, 1:01 PM

I'm working on an open-source system called SpinStep (just search GitHub) — it's a quaternion-driven traversal framework for orientation-based logic and spatial data structures.

It’s not directly tied to geolocation, but it could integrate nicely with something like Astradia. Since Astradia provides high-fidelity attitude data without relying on GNSS, SpinStep could use that orientation stream to drive autonomous behavior trees, scanning patterns, or state transitions — all without depending on coordinates or maps. Basically: orientation in, logic out.

Would love to hear from others thinking about orientation-first autonomy or mapless navigation.

by aredoxon 6/5/2025, 12:44 PM

Is there a similar system but much more affordable? I don't need it to be space-ready or milspec-grade or even compact, I was just thinking of something for tinkerers.

(The translation used by Sodern, "Star Tracker", gets me only camera gimbals to take astrophotography pictures)

Edit: found some leads around the cubesat community. Still to see if it works from the ground or in daytime.

https://ieeexplore.ieee.org/document/9179736

https://openstartracker.org/

by throw0101bon 6/5/2025, 12:55 PM

I'm curious to know why resurrecting (e)Loran seems to be a hard sell:

* https://en.wikipedia.org/wiki/Loran-C

China built out a GNSS backup with it:

* https://www.gpsworld.com/china-completes-national-eloran-net...

* https://rntfnd.org/2024/10/03/china-completes-national-elora...

* https://www.mdpi.com/2076-3417/13/23/12703

And there's been some rumblings from Korea and UK:

* https://www.gpsworld.com/south-korea-partners-with-broadcast...

* https://www.gpsworld.com/uk-leading-the-west-in-pnt-with-clo...

but no major moves in most countries, even though there's a recognition of GPS/GNSS vulnerabilities by even the (US) military:

* https://fedtechmagazine.com/article/2022/05/dod-transportati...

by throw0101don 6/5/2025, 1:11 PM

From four months ago, "Celestial Navigation for Drones":

* https://news.ycombinator.com/item?id=42767797

by Havocon 6/5/2025, 3:01 PM

Let’s hope they stick it on the airliners flying in the baltics

by hollerithon 6/5/2025, 2:03 PM

This could help North Korea and Iran develop ICBMs.

by contingencieson 6/5/2025, 12:17 PM

Use case = ?

3kg! That's a very high price to pay for reliability under GNSS denied environments in the day time. Surely you can get ~altitude from a barometer, ~position from inertial + IMU + prior position, and if you want more accuracy maybe DEM models + topography is better.

Therefore I'd suggest this is not going to be very useful on anything with a view toward efficiency (drones). This is more likely a 'quick fix' plugin for existing fat-plane avionics for 'identified risk' reasons. OK move commercially, but technically meh. The go to market plan would presumably be FUD around GNSS denied situations. Reality: unlikely to be problematic in most cases (known flight path, high altitude, low GNSS-denied environment dwell time).

by AStonesThrowon 6/5/2025, 11:26 AM

Wow! A daytime star tracker! What wavelengths does it use?

by blacklionon 6/5/2025, 11:25 AM

So, technology closed the circle. Nice.