It's official. Polaris will be part of this year's Google Summer of Code. For those of you who don't know what I'm talking about, Polaris is an open source project born after the last Open Source Cubesat Workshop held in Madrid. It was an idea which started to take form after the workgroup I proposed to discuss the opportunities and applications of machine learning onboard.
The end goal will be to develop a Python module that can take any set of satellite telemetry from the SatNOGS network, merge it with orbital information, and analyze it using different machine learning algorithms to diagnose potential issues. During the past weeks we have been defining the architecture together with the members of our Community. As you can see in the diagram below, the module will be split into three separate tools:
- Polaris-Fetch: get data from telemetry databases (such as SatNOGS) and prepare context information for machine learning algorithms.
- Polaris-Learn: feature engineering and selection along with machine learning algorithms to create prediction models, anomaly detection models and many other knowledge and situation awareness techniques.
- Polaris-Viz: data visualization focused on understanding what ML algorithms have output and why.
Now that I've introduced Polaris, it's time to welcome the student who will be working with us the next three months: Aditya Malshikhare. He's a 4th year undergrad from India who got into machine learning thanks to some interesting side projects. Apart from that, he also loves astronomy, photography and trekking (you can read more about his journey so far in his personal blog). Hugh Brown, Redouane Boumghar and me will be his mentors and we will be assisting him as much as we can.
In name of the Polaris mentors, I'd like to thank the Libre Space Foundation for helping us make this happen. The Foundation is doing a lot of interesting things, and they have just announced the next OSCW, which will be celebrated in Athens (Greece) the next October. We would also like to thank the rest of the students who applied for the project, there were a lot of interesting proposals but unfortunately there were not enough slots for everyone. We wish you the best and we invite you to stick around for contributions!