• We hope all of you have a great holiday season and an incredible New Year. Thanks so much for being part of the Space community!

Could gravity waves and AI help find solar system objects?

Status
Not open for further replies.
W

willpittenger

Guest
Suppose we have an AI system that simulates the solar system. It would run each orbit forward and backward comparing theoretical positions with those actually observed. Any anomalies are noted. Eventually, those anomalies will help the AI system predict the existence of a new comet, asteroid, or dwarf planet. (This would be the same way that Lowell predicted the existence of Neptune.)<br /><br />The system might then be tied into an automated telescope system that takes two pictures of the region at once from different vantage points. (You could settle for taking them days apart, but for that, you have to a rough idea of how far away the object is so you know how long to wait. Besides, taking both at once frees the telescope system up for its own search sooner.) Anything appearing above the background stars gets cataloged. (Then you would take a second set of pictures, but so you can calculate the orbit.)<br /><br />Each time you find something new, you rerun all the simulations. This makes new anomalies appear and you find new objects. It could be a continuous cycle. <div class="Discussion_UserSignature"> <hr style="margin-top:0.5em;margin-bottom:0.5em" />Will Pittenger<hr style="margin-top:0.5em;margin-bottom:0.5em" />Add this user box to your Wikipedia User Page to show your support for the SDC forums: <div style="margin-left:1em">{{User:Will Pittenger/User Boxes/Space.com Account}}</div> </div>
 
Q

qso1

Guest
Excellent suggestion. I would think it would be possible but I guess it depends on where AI is in its developmental stage at the moment. With all the hot Jupiters being discovered, I somethimes wonder if we don't have a hot Jupiter in a torch orbit that we cannot see because of proximity to the sun. What your proposing would be quite useful in finding such an object. <div class="Discussion_UserSignature"> <p><strong>My borrowed quote for the time being:</strong></p><p><em>There are three kinds of people in life. Those who make it happen, those who watch it happen...and those who do not know what happened.</em></p> </div>
 
M

MeteorWayne

Guest
Yer stretching things a bit there qso1 <img src="/images/icons/smile.gif" /><br /><br />With th amount of monitoring we do of the sun at mulitiple wavelengths, from multiple vantage points, we would have found anything orbiting close to the sun. <div class="Discussion_UserSignature"> <p><font color="#000080"><em><font color="#000000">But the Krell forgot one thing John. Monsters. Monsters from the Id.</font></em> </font></p><p><font color="#000080">I really, really, really, really miss the "first unread post" function</font><font color="#000080"> </font></p> </div>
 
H

heyscottie

Guest
Also, I think "AI" is misapplied here. All that is really required is huge amounts of computing power to run the extremely precise calculations needed.<br /><br />But I just don't know about finding objects like comets based on orbital deviations of planets and the sun. Certainly, an effect IS had on them, but I don't think we have the ability to make those measurements. In short, I think the noise in any of these measurements would be sure to completely swamp any signal from small objects.<br /><br />Also, it doesn't seem that you are actually using gravity waves in your calculations here, unless I am misunderstanding. You seem to be making positional observations and looking for deviations.
 
W

willpittenger

Guest
<blockquote><font class="small">In reply to:</font><hr /><p>Also, I think "AI" is misapplied here. All that is really required is huge amounts of computing power to run the extremely precise calculations needed.<p><hr /></p></p></blockquote><br />I figure that the analysis looking for those deviations might go faster with a AI neural net system.<br /><br /><blockquote><font class="small">In reply to:</font><hr /><p>Also, it doesn't seem that you are actually using gravity waves in your calculations here, unless I am misunderstanding. You seem to be making positional observations and looking for deviations.<p><hr /></p></p></blockquote><br />Gravity waves cause those deviations. <div class="Discussion_UserSignature"> <hr style="margin-top:0.5em;margin-bottom:0.5em" />Will Pittenger<hr style="margin-top:0.5em;margin-bottom:0.5em" />Add this user box to your Wikipedia User Page to show your support for the SDC forums: <div style="margin-left:1em">{{User:Will Pittenger/User Boxes/Space.com Account}}</div> </div>
 
H

heyscottie

Guest
willpittenger: I figure that the analysis looking for those deviations might go faster with a AI neural net system.<br /><br />Me: I'm not sure why it would. We are not doing complex pattern matching where we could train a nerual net properly. What we are doing is running deterministic differential equations forward with extremely high accuracy, and trying to find where reality deviates. That's just straightforward computing power. Earth-simulator like power.<br /><br />willpittenger: Gravity waves cause those deviations. <br /><br />I guess. But I won't be analyzing gravity waves. I'll be analyzing positions and velocities of observed bodies. You might as well say we are doing analysis of the electromagnetic waves, since they, after all, are what we see when we make those observations.
 
Status
Not open for further replies.

ASK THE COMMUNITY

TRENDING THREADS