What is Hive?
Hive is a collaborative research project aimed at improving cloud and rain detection by applying machine learning to CloudWatcher sensor readings and all-sky camera images. By aggregating data from many observatories in different locations and climates, we aim to build a detection model that outperforms threshold-based approaches in ambiguous conditions.
The findings will be published and will directly inform future improvements to CloudWatcher detection algorithms.
How to participate
Participants contribute the following data, shared via the Hive project website:
- Hourly CloudWatcher sensor readings
- One all-sky camera image per hour, synchronised with the sensor data
- GPS coordinates and altitude of the observatory (kept confidential — location data is used only to group readings by climate zone)
To join Hive or for more information, visit the Hive project website or contact info@lunaticoastro.com.
What Lunatico commits to
- All findings will be publicly disclosed
- Improved cloud detection systems based on the research outcomes
- Participants are invited to help characterize sky imagery via the web interface (cataloguing images as clear / cloudy / rainy)
Project timeline
| Period | Milestone |
|---|---|
| July 2021 | Project launch; initial participant contact |
| August 2021 | Cloud server setup |
| September 2021 | Individual participant assessment |
| October 2021 | Data collection begins |
| August 2022 | Web interface for image cataloguing |
| Ongoing | Data collection and analysis |
