5
$\begingroup$

edit: Possibly helpful information in Gizmodo and YouTube.

In general spacecraft are supplied with propulsion systems often with substantial redundancy in the plumbing and engines or thrusters so they can dependably and reliably perform orbital maneuvers in space. The emphasis is on the word perform. They wait for commands from Earth before doing anything with it. You could say that spacecraft are carefully micro-managed.

For deep space missions, this is done with an abundance of caution to protect the substantial investment of time, people, and resources and the science to be gained. For Earth orbit, in addition to protecting all of those (with commercial gain replacing the scientific gain in most cases) there is the additional responsibility of not interfering with all of the other satellites in orbit.

I was surprised that even the very expensive James Web Space Telescope does not seem to have the ability stay put by itself as hypothesized in this answer, and without constant station-keeping instructions successfully received from Earth, would just drift away from it's libration point orbit.

To contrast, planetary explorers (as opposed to spacecraft) such as rovers and in the future flyers and/or hoppers execute sequences of tasks in rapid succession that require first-hand knowledge of the local environment which makes it beneficial to operate semi-autonomously. For example, based on experience from a number of past and present rovers on Mars, the Curiosity rover on Mars is capable of keeping itself busy for a day or so both in movement and sample taking and analyzing samples as described in this answer, and the EXOMars rover will potentially have a degree of autonomy as well.

So my question is, What are the future prospects for spacecraft autonomy?

$\endgroup$
  • $\begingroup$ The fact that this question may have an answer here is not a coincidence. I felt that that answer needed its own question specifically about spacecraft autonomy. The other question is only about technology for determining the direction of a delta-v maneuver. $\endgroup$ – uhoh Dec 3 '16 at 0:51
  • 3
    $\begingroup$ Experiences to date have not been overwhelmingly successful. Which is to be expected with experimental technology. nasa.gov/pdf/148072main_DART_mishap_overview.pdf $\endgroup$ – Organic Marble Dec 3 '16 at 3:57
  • $\begingroup$ The proposed [helicopter drone] for Mars 2020 would need to be mostly autonomous. It’s flight will be on the order of ~2 minutes duration so it will need to be able to fly, identify a landing site and land without any input from ground control which we be tens of lightminutes away. Would that be a valid example of a prospect? $\endgroup$ – Jack Jun 30 '18 at 11:45
  • $\begingroup$ Forgot the helicopter drone link $\endgroup$ – Jack Jun 30 '18 at 12:07
  • $\begingroup$ @Jack in this case I can't say much about an answer that hasn't been posted yet. Since I've already said "...rovers and in the future flyers and/or hoppers execute sequences of tasks in rapid succession that require first-hand knowledge of the local environment which makes it beneficial to operate semi-autonomously", yes this is one potential example. Can you think a way of extending this into an answer about future prospects? I admit this is a slightly soft question. If you could add one or two more future examples, that might cover the "prospects" angle better. $\endgroup$ – uhoh Jul 1 '18 at 5:40
1
+200
$\begingroup$

Spacecraft autonomy has always played an important role in the exploration of the solar system and beyond and our reliance on increasingly independent spacecraft will only grow. There will be varying levels of autonomy depending on the task at hand. Immediate applications of increased autonomy could possibly be onboard anomaly detection, health and resource management. Followed by more complex decision making and guidance, navigation & controls.

I foresee the near-future of spacecraft autonomy being more model-driven approaches, as it's historically been, as opposed to data-driven models like deep learning. There are two reasonings for that: 1. Deep learning is very compute intensive, and often spacecrafts are highly computationally resource constrained. 2. Deep learning lacks explainability. A neural network is considered a black box, and if we want verification & validation of the algorithms, we need to be able to see into this black box.

As access to space continues to be democratized, I see a possibility where CubeSats could actually be the propeller of more novel data-driven autonomy algorithms. The most recent MarCo CubeSat missions to Mars could entice the industry to send more of these spacecrafts. And as a result, take more risks in the type of autonomy is deployed. The other side to this coin is that physics-based simulation engines continue to get better and better. This will allow researchers to apply creative autonomy techniques to highly-realistic environments. Not only that, there is a lot of research going on in the area of synthetic data. This could be highly beneficial to generate high quality data that a rover's deep neural net could be trained on prior to being deployed. Check out this paper on using deep reinforcement learning for autonomous imaging and mapping of small bodies.

Also this paper on using reinforcement learning for motion planning of hopping rovers.

Check out this great survey on AI trends in GN&C

Conclusion:

There are a lot of great opportunities for spacecraft autonomy and a lot of factors that come into play when deploying increasingly complex algorithms. Ultimately, the future of spacecraft autonomy looks very bright.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.