Hacker News Reader

Comment by Waterluvian | Hacker News Reader

I just had a daydream of Satellites as a Service. "give us a docker image extending our base image, no larger than x MB delta, and we will run it on a chosen satellite for a certain time slice for $y"

I can see that future. Even if purely as a learning tool for students.

View on HN

Would the satellites have some known set of sensors or something like that? Just trying to figure out what would be the point, other than just the coolness of running code in spaaaaaaaace.

Yes, some common API for access to attitude sensors. Maybe a radio. But honestly the point I envision is that it's cool. And teaches kids/teens that this space stuff isn't that far out of reach.

I just remember growing up thinking that I'm a C student and so all this stuff is to be left to the smart people. And that error in perspective drove my thinking for a long time.

You never really know what will be the spark that helps someone discover their passion. Maybe it's Nyancat Sputnik.

Raspberry Pi In The Sky

Students can design, build, validate and launch their own cubesats through special programmmes.

Even cooler if one of them launches a flying data center. Limited bandwidth would be an issue.

Bandwidth is usually less of a problem than latency.

Check out the AstroPi challenge [1] for a similar idea on the ISS. Including support for multiple sensor and cameras.

[1] https://astro-pi.org/

AFAIK, that's what Satellogic (https://www.satellogic.com) sells. You may want to contact them.

Disclaimer: while I am not directly involved with what they do, I happen to know personally both of the founders and a lot of the people working there.

Considering the cost of storage it's probably cheaper to launch N sensors and just keep a record raw everything. Which is more or less the model of the existing earth-sensing satellite world.

That defeats the whole point by optimizing for the wrong thing.

Reply on news.ycombinator.com