we need traffic lights for satellites

[madires]

ESA had to perform a collision avoidance manoeuvre for Aeolus because of SpaceX' Starlink:
https://twitter.com/esaoperations/status/1168533241873260544

SpaceX Refused To Move A Starlink Satellite At Risk Of Collision With A European Satellite:
https://www.forbes.com/sites/jonathanocallaghan/2019/09/02/spacex-refused-to-move-a-starlink-satellite-at-risk-of-collision-with-a-european-satellite/#6e2c36d51f62

Aeolus is a little bit longer in space than the Starlink satellite. So we can guess what will happen with tons of LEO internet satellites and SpaceX' attitude.

[Ysjoelfir]

Naaaah, just build a three dimensional matrix of hyperloopses in space for every sattelite out there. Problem solved. It will only cost 1$ for a sat to travel through the loop!

[bd139]

Raytheon have a decent solution for this in development. It's called the MOKV, the Multiple Object Kill Vehicle 

[nuclearcat]

As i understand they have quite weak Krypton thruster, and it can operate only some time on each turn (90min), i guess only when panels are insolated.
They have single panel, again according "internet sources" approx 3kW, thruster according to publicly available data on efficiency can provide approx 150 millinewtons.
Weight of satellite is approx 227kg.
Current orbit according to article 320km.
It means 7718.27m/s current orbital speed. Just if they raise orbit 10km they will have orbital speed 7712.51m/s, and need ~24h of continuous engine operation, taking in assumption ESA tweet "The manoeuvre took place about 1/2 an orbit before the potential collision. "
I have some doubts that we can say these satellites have proper collision avoidance capabilities.
Does anyone have more info?

To test my theory i took this data on their orbit change maneuer.

According to news orbit raise maneuver was (very approximately) from 2019/05/27 to 2019/06/14, 19 days. If we assume data above is correct, they need to run at least 56% of time such thruster.
On the orbit 440km satellite will spend in shadow (keeping in mind atmosphere attenuation too) t=τ*(2θ)/(2π)m where is τ = orbital period θ = arcsin(r/(r+h)) (r is 6371 + 100) approx 39% of all time.
It looks like close to true...

Knowing how I am prone to make horrible mistakes in numbers and how in a hurry i wrote this assumption - maybe I wrote nonsense above

[schmitt trigger]

Raytheon have a decent solution for this in development. It's called the MOKV, the Multiple Object Kill Vehicle 

Which, instead of a single object one would have to watch for, would generate thousands of fragments creating a lethal shrapnel cloud orbiting the earth.

As a matter of fact the Chinese military performed such a test a few year ago, and created such an orbiting hazard.

[magic]

Yeah, cheaper to just nuke SpaceX and let their satellites safely fall down into the atmosphere

[SiliconWizard]

Oh, and who said it would never get crowded enough up there that we should care?
 

[Red Squirrel]

Will be interesting once SpaceX does in fact launch all the sats they plan to.  If they do it I'm sure another party will want to compete and do so as well.  it's going to be like the 401 up there except instead of travelling at 0.01m/s you're travelling at 7000m/second.

[madires]

SpaceX satellite was on “collision course” until ESA satellite was re-routed:
https://arstechnica.com/information-technology/2019/09/spacex-satellite-was-on-collision-course-until-esa-satellite-was-re-routed/

SpaceX underestimated the risk of collision and then a bug in the paging system prevented them to receive the US Air Force's update about the increased collision probability. A convenient excuse?

[Psi]

Traffic lights wouldn't work, but i'm sure you know that.

But maybe a TCAS system like planes use could work.
Where the satellites can take their own action to move out of the way in opposite directions if they sense a proximity issue.

It would likely have to be based on a global tracking system of all sats, rather a than radio signal/transponder system like TCAS.
Depending on the orbits the collision could be a slow approach or a head on collision at km's per second with little time to detect a short range transponder before impact.
And the issue of satellites with hall/plasma/ion thrusters not being able to move fast means you need lots of advanced warning.

[timgiles]

I still find it hard to believe in a three dimensional space (spherical skin with depth of x meters or kms) the size of the earths diameter plus a tiny bit - that satellites really get that close to each other - ever! Unless they are programmed to do so (conspiracy theory time!).

Am I wrong? Please help my old grey matter understand if so. I just dont see it happening. Sure, space debris, but that is many many magnitudes higher density.

[Brumby]

Every orbit of every satellite is different.  They will have different apogee and perigee and they will drift.  The more satellites you have, the longer they are circling the Earth, the greater the chances that you will have an intersect - and destruction.

The reason there hasn't been a big problem with collisions is because of the volume of space involved, but as the number of satellites increase, these odds get less.

It is the same sort of problem we have had to deal with in regards to atmospheric pollution.... When it first started out, the impact on the environment was ignorable - so it was ignored.  Then it got worse and now we are facing major problems.  The only reason satellites have been given the attention they have is because of the cost.

[Psi]

Also, when they say collision it's not a certainty it's a statistical probability and it's very small.
When your satellite is worth 100 of millions in replacement costs even a remote chance of a collision is too high.

[madires]

But maybe a TCAS system like planes use could work.
Where the satellites can take their own action to move out of the way in opposite directions if they sense a proximity issue.

It would likely have to be based on a global tracking system of all sats, rather a than radio signal/transponder system like TCAS.
Depending on the orbits the collision could be a slow approach or a head on collision at km's per second with little time to detect a short range transponder before impact.
And the issue of satellites with hall/plasma/ion thrusters not being able to move fast means you need lots of advanced warning.

We already have a tracking system of all satellites and larger debris. TCAS for sats sounds neat, but I think it wouldn't be feasable. Current sats in space can't be upgraded, and we still have to track debris. What about rogue sats or ones with a broken TCAS?

[Psi]

But maybe a TCAS system like planes use could work.
Where the satellites can take their own action to move out of the way in opposite directions if they sense a proximity issue.

It would likely have to be based on a global tracking system of all sats, rather a than radio signal/transponder system like TCAS.
Depending on the orbits the collision could be a slow approach or a head on collision at km's per second with little time to detect a short range transponder before impact.
And the issue of satellites with hall/plasma/ion thrusters not being able to move fast means you need lots of advanced warning.

We already have a tracking system of all satellites and larger debris. TCAS for sats sounds neat, but I think it wouldn't be feasable. Current sats in space can't be upgraded, and we still have to track debris. What about rogue sats or ones with a broken TCAS?

It would be for new sats, It would obviously not help older sats hitting other older sats.
But any new sats could avoid any olders sats or debry using data from ground based sat tracking.
Newer sats could also coordinate directly with other newer sats to avoid each other

[SparkyFX]

But maybe a TCAS system like planes use could work.
Where the satellites can take their own action to move out of the way in opposite directions if they sense a proximity issue.

The orbital mechanics require satellites to take action while on the other side of the orbit the potential collision would occur, at least that is the location where you get the biggest change of height in the shortest time with the lowest amount of fuel.
You could move a satellite sideways (change its inclination), but it would cost lots of fuel for very little change.

So there is no direct line of sight/communication and for proximity based systems this would be too late, as most satellite thrusters do not have a lot of thrust (provide only a small fraction of the weight of the satellite, making this a very slow process). They are built for small corrections that need to take place at the right time.

As far as i understood it, TCAS in planes more or less assumes all others stay on their course on their altitude (as in a circular orbit), satellites do not necessarily have to have a circular orbit, making the calculations and detection way more complex.

Video about collisions in space and collision avoidance:

[nuclearcat]

But maybe a TCAS system like planes use could work.
Where the satellites can take their own action to move out of the way in opposite directions if they sense a proximity issue.

The orbital mechanics require satellites to take action while on the other side of the orbit the potential collision would occur, at least that is the location where you get the biggest change of height in the shortest time with the lowest amount of fuel.
You could move a satellite sideways (change its inclination), but it would cost lots of fuel for very little change.

So there is no direct line of sight/communication and for proximity based systems this would be too late, as most satellite thrusters do not have a lot of thrust (provide only a small fraction of the weight of the satellite, making this a very slow process). They are built for small corrections that need to take place at the right time.

As far as i understood it, TCAS in planes more or less assumes all others stay on their course on their altitude (as in a circular orbit), satellites do not necessarily have to have a circular orbit, making the calculations and detection way more complex.

Video about collisions in space and collision avoidance:

I think it is quite simple, if satellites able to get their position anyhow (GPS on orbit?). They need few points and its enough to build TLE. Send it to ground stations across the globe, and they will assist this "satellite-TCAS", same as flight dispatchers.
The only difference is that the satellite is better to start the maneuver in advance. For example, Starlink is very slow to move.
Dispatchers can also help to plan orbit movements.

[klunkerbus]

I seem to recall that the Kessler Syndrome from the 70's suggested it was just a matter of time before LEO became a disaster zone.

[Mr. Scram]

I still find it hard to believe in a three dimensional space (spherical skin with depth of x meters or kms) the size of the earths diameter plus a tiny bit - that satellites really get that close to each other - ever! Unless they are programmed to do so (conspiracy theory time!).

Am I wrong? Please help my old grey matter understand if so. I just dont see it happening. Sure, space debris, but that is many many magnitudes higher density.

Aviation safety used to depend on exactly that reasoning. Then flying became more popular and people started dying.

[m98]

There really should be some kind of space traffic control a commercial satellite operator should have to stay in touch with to coordinate maneuvering.

[coppercone2]

lol, pay protection money for your satellite.

[Red Squirrel]

Is a satellite considered a vehicle, wonder if you can get car insurance for one. 

"Sir, can you clarify your speed before the collision?  Are you still on the accident scene? We're going to need to send an adjuster. "

There really should be some kind of space traffic control a commercial satellite operator should have to stay in touch with to coordinate maneuvering.

This is probably the best bet, I think they already do track space objects including junk, so it could be the thing of predicting collisions way ahead of time, calculating the most efficient maneuver, and time, and then sending the info to the operator.  Kinda like a "btw your satellite will collide with xyz on this date, to correct, follow this maneuver".

[coppercone2]

are they gonna have to pay NORAD for all the extra shit they put up there? should there be a rule against American companies using foreign nation to launch satellites under their control? should a space control cabal be formed?

[windsmurf]

Is a satellite considered a vehicle, wonder if you can get car insurance for one. 

"Sir, can you clarify your speed before the collision?  Are you still on the accident scene? We're going to need to send an adjuster. "

There really should be some kind of space traffic control a commercial satellite operator should have to stay in touch with to coordinate maneuvering.

This is probably the best bet, I think they already do track space objects including junk, so it could be the thing of predicting collisions way ahead of time, calculating the most efficient maneuver, and time, and then sending the info to the operator.  Kinda like a "btw your satellite will collide with xyz on this date, to correct, follow this maneuver".

http://www.parabolicarc.com/2019/08/01/satellite-insurance-rates-increasing-after-failures-of-vega-worldview-4/
https://en.wikipedia.org/wiki/Satellite_insurance

[lilstevie]

This is probably the best bet, I think they already do track space objects including junk, so it could be the thing of predicting collisions way ahead of time, calculating the most efficient maneuver, and time, and then sending the info to the operator.  Kinda like a "btw your satellite will collide with xyz on this date, to correct, follow this maneuver".

The problem is there's a margin of error in the calculations, like with everything the real world is far from perfect, without mapping the exact amount of gravity and resistance from the upper layers of the atmosphere for every mm^2, we won't know exactly where the satellite will end up on its track. That's why this story, and collision tracking in general talks in probability of a collision. The collision was in no way guaranteed, it was just too much of a risk to chance it.