Inspiration
The idea of a new world made us realise the new world which is right in front of us, which we have never explored. Of course, that being space. Essentially, we wanted to use the untapped potential of mining asteroids to help extract valuable resources from space.
What it does
eta. is an interactive map of the asteroid belt that works out where to mine and where to place stations, so we can make things in space instead of launching them from Earth. Essentially, you pick a product such as steel, and Eta finds the asteroids holding its ingredients/ores and places stations optimally next to them.
How we built it
We started from IMCCE's catalogue of over 1.5 million asteroids and kept the 65,128 with enough data to estimate what they are made of. We then linked ores to products through a network of 322 processes and used weighted k-means to place stations. The planner runs in Python and the map in Next.js with WebGL.
Challenges we ran into
The biggest challenge we faced was implementing the correct algorithm. We spent multiple hours figuring out what would be the most correct and useful algorithm in our case, one that followed most of the requirements we needed, such as being able to put stations close to certain ore deposits for efficient routing. We tried many different algorithms until we ended up with the one that fit our requirements specifically and with an optimal/fast solution
Accomplishments that we're proud of
We are really proud of mapping over 65,000 real asteroids and rendering over 200 distinct stations optimally, with more than 16,000 across a full production chain in about 3 seconds. We also ran the placement 1,000 times with different compositions, and the stations barely moved, which showed us our results hold up.
What we learned
We learnt a core skill of teamwork by learning how to delegate tasks as a team, how to essentially work in parallel efficiently using git control. We learned to do this by having a clear set of ideas, which mapped out exactly what each person in the team is working on and allows concurrency. A main aspect that we learnt is that having a clear idea is very important, as it makes implementation a lot smoother, rather than having a rough idea and having a messy set of tasks shared between everyone.
What's next for Eta
Next, we want to let you plan for several products at once, with the stations updating automatically. We also want to cover all 1.5 million known asteroids as more get classified, and add real orbital mechanics and mining rates so Eta can plan how a space industry could actually run.
