4 ms·
I would imagine the problem with this model is that there isn't actually a whole lot you can do if you rent time on a satellite. Given an orbiting sensor suite,
by mechagodzilla 4y ago
I would imagine the problem with this model is that there isn't actually a whole lot you can do if you rent time on a satellite. Given an orbiting sensor suite, all you can do is log the sensor data and send it back to earth - you can't really run experiments or anything. So the company that owns the satellites can do that, and then try to . . . sell the data? That's basically the business model. That works for satellite imaging companies, but what other sensors might that work for? And how many paying customers could you reasonably have?
- d_silin 4y agoYou can run ML algorithms on imagery in space, and do a bit more than just downlink sensor data. There is already one satellite that works as such an experimental platform: https://www.groundstation.space/esa-funds-12-new-experiment-to-fly-on-ops-sat/ https://www.groundstation.space/esa-funds-12-new-experiment-...