Debris DeOrbiters

Text from chat about "world changing":

miguel.delcastillo

7:49 AM

Good luck! Sorry for the confusion.

If you find a way to forecast possible debris collisions through ML, that would be outstanding and world-changing! But it's not a requirement.

Space-Track has conjunction (i.e. close calls) datasets. That could be your training data.

Search for "space-track conjunctions" for information on the data, and "space-track API" for information on how to acquire it.

And if you tie all of this to a GIS visualization that shows the entire cloud of space debris, then you're not only complying with this challenge's requirement, the project would most assuredly be the winner.

I saw in chat that they removed the AI/ML requirement from the Space Debris challenge. I’m going to continue with my idea, anyway. It was mention in chat that if AI/ML could be used to predict collisions, it would be “world changing.”

Proposed Project Title: Using deep learning to locate space debris

ST9 Solar Sail - A Potential Meteoroid and Orbital Debris Sensor  (pg. 6-7)

https://orbitaldebris.jsc.nasa.gov/quarterly-news/pdfs/odqnv9i1.pdf