Idea: an Open Source Spacecraft Compute Platform

Troy · 2021-02-05T06:25:26+00:00

So here’s a crazy idea. CC @mjdsmith965 What if: we try to pull together a community-driven, open-source spacecraft/satellite bus system we open access to...

Idea: an Open Source Spacecraft Compute Platform

Troy

So here’s a crazy idea. CC @mjdsmith965

What if:

we try to pull together a community-driven, open-source spacecraft/satellite bus system
we open access to the bus by developing it into a cloud solution so people around the world can remotely connect novel payloads to it and run applications on it - as if it was a cloud service

There are a number of open-sourced CubeSats around, but I feel that they aren’t quite good enough to be a more powerful platform. I’m thinking a spacecraft bus for craft in the 100-200kg range. Could be useful for telecoms sats, EO sats, or even spacetugs.

Moonshot would be happy to host the hardware. I think we could start to see some incredible innovations if that was made available, and its exactly this kind of open activity that I build Moonshot to encourage.

It would be good to check out FOS licences that are more appropriate for hardware, which are only just coming to up, like this one.

What do we reckon? 🤔

rocketgirlmars

Troy Awesome idea !

AndrewButtery

Hi Troy,

Still trying to get my head around this. Are you propsing to:

Establish a new standard satellite bus and open source it.
Then people can remotely create “virtual” payloads based on the bus. Either using a predefined set of plug and play component parts for common requirements (e.g. comms) or custom parts they define for non-standard requirements.
They can then either build their own (which should be cheaper as it’s now based on a standard) OR have the option to have the payload physically fabricated, tested and launched with a few click of a button (Satellite build as a service) Thinking this is not the communities role here, but that it may provide an oppurtunity? Sort of like designing PCB, sending them off to a fabricator and getting the boards back in the mail. Instead, you’d get a satellite in space…
For those that couldn’t afford/didn’t need a custom payload that the community would (somehow) be able to get their own satellite up based on the standard where people can run their own various experiements based on sensors/compute…

Or have I completely missed the mark here?

Troy

Yes, exactly. The cubesat, for example, is barely a standard. It’s great, but there’s so much variation in how everyone implements it. Further, it might be fine for shoe-box sized satellites for small experiments, but the most valuable satellites are all going to be in that 100-200kg range for the foreseeable future (check out the size of Starlink for example).

At the moment there are propriety satellite busses that teams can design for, but I’m imagining:

an open-source satellite bus design that anyone can manufacture themselves, allowing teams to focus on the function of the payloads rather than reinventing the wheel or needing to rely on a large aerospace partner who won’t give you the time of day without a multi-million dollar budget ready for them on day 1 of discussions.
The ability to have a few of those satellite busses made available to people via the internet, in the same way, AWS makes available various computers and devices via the cloud so that people can easily spin up a ‘virtual satellite’ to develop peripheral hardware and software on top of without needing to first invest in building a 100-200kg satellite.

I reckon that these two things would create a secondary boom in space infrastructure projects, because it would drastically reduce the barriers to entry even further than the cubesat did.