In space, it’s very important not to run into anything. This sounds obvious, but it’s worse than you think. Because it can cost thousands of dollars per pound of payload/spacecraft weight to launch something into orbit, spacecraft are often made of very very thin materials. Thin means easily broken, and a tiny hole means all your oxygen (and pressure!) goes flying out, probably explosively given the pressure differentials. Now you get to find out for sure how you die in space.
(Spoiler: all the ways are terrible, avoid this!)
In addition, because there is no drag from air resistance, a very small object that goes flying off (like a dropped screw, or a piece of a decomissioned satellite that has broken up) will keep whatever its initial velocity was. Orbital velocities are necessarily very large, in order to keep the object in orbit, with higher velocities the closer you are to the earth.
The ISS orbits the earth at 28,165 kilometers per hour, so a 1 gram screw that fell off the ISS might hit something with a force of
F = mv2 (so with some unit conversions)
= .001kg ∗ (7823.61m/s)2
= 78.23 Newtons.
If you aren’t familiar with what a Newton feels like, that’s like having 7 kilograms fall on you.
Not a nice feeling!
And there are a lot of pieces of space junk out there. The DoD’s Space Surveillance Network tracks over 27,000 pieces of orbital debris, most of which are softball sized or larger. They estimate 100 million pieces in the millimeter size range, just in the most useful part of Earth’s orbit for human space systems. So with all that “junk” out there, a meteorite/space junk identification system whose job is identifying pieces of “stuff” that might be on a collision course with your spacecraft is pretty important. The operator needs to know with a pretty high certainty if an object is going to encounter the spacecraft, and how soon.
If an automated system, like the Lockheed-Martin Space Fence, is used to flag potential threats, signal detection can help by setting the automatic system’s criterion at an appropriate place to balance the threat from potential impacts against the time demands on the user to investigate false alarms. Even once a potential threat has been identified, the system needs to alert the user about the threat and they need to analyze it. A display for identifying micrometeorite threats needs to have a clear and uncrowded display to avoid misidentification - perhaps in the list of all tracked items, the potential threats could be identified in bold red while the non-flagged items are in faint grey. This will improve the user’s sensitivity to detecting the output of the automation, which will allow them to more quickly evaluate the threat and prepare for it if needed.
Here's a video overview of how signal detection/object tracking works in the Space Fence system:
This is a really serious topic and for some reason I am laughing at it. Thanks for the insight and the laughs! In all seriousness, this is super interesting and super important and just goes to show that signal detection needs to be apart of every human machine system. I wonder how all the space debris will be dealt with in the future...there has to be some limit to the amount of stuff that can surround the Earth. It will be interesting to see how its dealt with.
Hi Kat, I found this topic very interesting and I think you did a very good job explaining it. I was always curious how they would detect potential threats as the cost of an accident in space would be really high. If they set a relatively high criterion, they might overlook potential threats; on the other hand, if the criterion is too low, then the system might have to alert the users constantly which might create confusion and bad user experience. I wonder how they are going to have a balanced criterion for both extreme situations. Also thanks for including the videos!