Transform a consumer smart telescope into a fully autonomous scientific instrument for variable star photometry.
SeeVar is an autonomous observing and data pipeline for the ZWO Seestar S30-Pro. Its purpose is simple:
- plan scientifically useful observations
- capture trustworthy raw science frames
- reduce them into defensible photometry
- prepare results for AAVSO submission
This project no longer treats the telescope as a consumer imaging toy. It treats it as a small robotic observatory.
SeeVar is built for long-term monitoring of variable stars, with current emphasis on:
- Long Period Variables: Mira and Semi-Regular stars
- Cataclysmic Variables during outburst and follow-up
The guiding goal is not pretty pictures. It is repeatable, honest measurement.
Observation cadence follows AAVSO-style scientific needs rather than casual imaging habits.
SeeVar is in active beta and scientific hardening.
As of April 2026, the project has already confirmed:
- direct hardware control via ZWO's built-in ASCOM Alpaca driver on port
32323 - autonomous nightly planning
- sovereign flight execution with a canonical
A1-A12sequence - Bayer-aware raw FITS capture
- simulator support for end-to-end workflow testing
- a postflight architecture now frozen around a canonical
P1-P8chain
The current chapter is v1.9.x: postflight scientific hardening.
That means the next priority is not discovering more hardware control. It is making the science chain more defensible:
- real solved WCS in postflight
- dark-calibrated working frames
- sigma-clipped comparison-star ensembles
- deterministic AAVSO report staging
Small telescopes can do real science if they are operated consistently.
The Seestar S30-Pro is affordable and capable hardware, but its consumer workflow is not designed for disciplined photometry. SeeVar exists to bridge that gap by adding:
- deterministic mission planning
- scientific cadence awareness
- raw FITS custody
- postflight quality control
- observatory-style logging and state tracking
The result is a telescope that can work unattended while still producing scientifically meaningful output.
SeeVar is organized as a sovereign observing pipeline:
-
Preflight Builds the nightly plan, applies cadence and horizon logic, checks weather and hardware, and decides whether the mission is allowed to start.
-
Flight Executes one canonical target sequence per object using the
A1-A12flight chain: target lock, safety gate, session init, slew, verify, settle, exposure planning, acquire, quality gate, and commit. -
Postflight Processes captured frames using the
P1-P8science chain: ingest, calibration matching, calibration application, astrometric solve, source measurement, ensemble calibration, quality verdict, and commit/report. -
Oversight Dashboard, logs, notifier, and ledger state remain available throughout the entire mission.
SeeVar communicates with the Seestar through the official Alpaca REST interface exposed by the telescope firmware.
Confirmed device access includes:
- Telescope
- Telephoto camera
- Wide-angle camera
- Filter wheel
- Focuser
- Dew-heater switch
This means:
- no phone app required
- no session master lock
- no middleware required for the core control path
The telescope is treated as a directly controlled instrument.
SeeVar captures and preserves raw science FITS. Flight ends when a trustworthy raw frame has been captured and committed.
SeeVar does not rely on naive debayering for production photometry. Its scientific direction is Bayer-aware source measurement directly on the sensor mosaic.
Current reporting direction:
- science channel:
G - reporting code:
TG
A magnitude is only trustworthy if the pipeline can justify:
- detector truth
- astrometric truth
- photometric truth
That is why postflight is now being hardened aggressively.
Scientific data should not depend on SD-card luck.
Recommended deployment:
- Raspberry Pi running Debian Bookworm
- external USB storage
- mirrored RAID1 data storage for observation products
- live volatile state in RAM where appropriate
The operating system lives on the SD card. Observation data, caches, and science products should live on more reliable storage.
Install on a fresh Raspberry Pi OS Lite 64-bit system:
bash <(curl -fsSL https://raw.githubusercontent.com/edjuh/seevar/main/bootstrap.sh)
The bootstrap process is intended to:
install dependencies
create the Python environment
collect site and telescope configuration
prepare system services
bring the observatory into a runnable state
For full instructions, see INSTALL.md.
For upgrading an existing checkout, see `UPGRADE.MD` or run:
```bash
cd ~/seevar
curl -fsSL https://raw.githubusercontent.com/edjuh/seevar/main/upgrade.sh | bashDocumentation Project doctrine and architecture live in the logic documents under dev/logic/.
Good starting points:
CORE.MD ARCHITECTURE_OVERVIEW.MD STATE_MACHINE.MD FLIGHT.MD POSTFLIGHT.MD PHOTOMETRICS.MD ROADMAP.md Astropy SeeVar contains some custom implementations that grew out of the project's early phases, before the full breadth of Astropy was properly appreciated.
Current direction:
use Astropy more where it improves correctness, maintainability, and scientific trust keep custom code where the problem is genuinely SeeVar-specific: Bayer-aware photometry Seestar-specific hardware behavior mission-state orchestration custody and observatory workflow This is an area of active review, not a philosophical rejection of Astropy.
Beta Expectations SeeVar is not pretending to be finished.
What is already real:
Alpaca control nightly planning simulator-supported mission flow raw FITS capture postflight scientific doctrine What is still under active hardening:
real solved WCS as a hard postflight dependency dark-calibrated science frames ensemble sigma clipping final reporting path That is the honest state of the project.
Contributing Testers and technically minded contributors are welcome.
Please open an issue first if the change affects:
mission sequencing protocol assumptions scientific validity ledger semantics observatory doctrine For contribution standards, see CONTRIBUTING.md.
Philosophy Good hardware deserves serious use.
A small telescope, careful automation, and scientific discipline can produce real observations night after night. SeeVar exists to make that possible without pretending that autonomy is the same thing as trust.
The project’s rule is simple:
If a frame is not proven, it is not accepted.