What would be the most challenging aspects of a reusable satellite scheme similar to the one shown in this video? For the kilo-constellations (e.g. SpaceX's StarLink with 4,500 to 7,500 satellites) the video proposes a more orderly way to deorbit and recover satellites at EOM. In-air recovery reduces the possibility of unsightly junk on the ground and on people's roofs, and offers the possibility of reusing at least some aspects of the satellites.

Don't laugh, reusing first stages used to be unrealistic as well.

I'm not asking if you would recommend doing this, or if you think this would be a good idea. Just what are the most challenging aspects?

Transcription of the video mentioned above and linked below, cued at 08:54 which can be verified by its closed captions:

Twenty satellites burning up a week, or one thousand per year is insane. According to SpaceX simulations, there is also a high chance of a couple kilograms of metal surviving reentry, and these pieces will reenter and crash into the ground somewhere!

Of course most will do so in the ocean, but some will inevitably crash into the ground. They could harm people, and even cause wildfires.

Of course, so could Not A Flamethrower.

  • 5
    $\begingroup$ I initially misread "starlink" as "starbucks" in the first video. That it's a light green circle and that it's the exact shape of a coffee cup didn't help. $\endgroup$ Apr 28, 2019 at 2:44
  • $\begingroup$ @JohnDvorak Clearly a case of evolutionary convergence; no other explanation is even remotely possible ;-) $\endgroup$
    – uhoh
    Apr 28, 2019 at 2:55
  • 4
    $\begingroup$ As for the drone sling - it sure looks super-cool, but can you actually predict the reentry of satellites with enough precision to get there in time? SpaceX's first stages have trajectories planned well ahead and controlled with computers. Or will these satellites have enough brains to call a sling for themselves and then dive into it when one is available? (Also, programming the drone AI to brace themselves for impact and not smack into each other will be fun) $\endgroup$ Apr 28, 2019 at 6:11
  • $\begingroup$ @JohnDvorak I wonder if you could develop this into an answer; I've asked for the most challenging aspects and that seems to be what you are coming up with... $\endgroup$
    – uhoh
    Apr 28, 2019 at 7:24
  • 1
    $\begingroup$ As it's written it is your own proposal. If it's not your proposal, then a reference to someone proposing it is missing. $\endgroup$
    – asdfex
    Aug 15, 2020 at 13:38

1 Answer 1


I imagine the most difficult hurdle to overcome in a satellite recovery program like you describe would be the financial one, as there is very little incentive for SpaceX (or any company) to do so.

First though, I think you're overestimating the problems/dangers of reentering space debris. Let me outline my reasoning:

  • Any debris which survives reentry will statistically speaking land in the ocean and disappear forever. Yes, initially, some debris might come down over land, but with the sheer amount of reentry events occurring, I'd expect SpaceX's predictive modeling of reentry to improve by leaps and bounds. Since Starlink satellites do have propulsion systems, adjusting their trajectory so that surviving material never hits land should be possible.
  • From an environmental standpoint, satellite debris is completely negligible. Even if a metric ton or two of (rather pure) metal returns to Earth and splashes down in oceans every year from deorbited satellites, it would have less an environmental impact than a single person driving their car into a lake or an ocean globally every year.
  • As for wildfires, I'm still of the opinion that all (if not an overwhelming majority) of the debris can be redirected into the ocean. Regardless though, lightning strikes or negligent humans are almost infinitely more likely to cause a wildfire than a hot chunk of metal (which might bury itself before it causes a fire).

Secondly, I don't think that satellites are expensive enough to justify reuse:

  • Starlink satellites are expected to have lifespans of around 5 years, and this limit comes not only from fuel depletion and hardware degradation, but also from technological obsolescence. All of the currently deployed Starlink satellites are without their inter-satellite laser links and can only serve a limited amount of customers at a time. Starlink will doubtlessly want to expand in the future, and eventually a 5-y/o satellite will be so behind in capability when compared to newer ones that it's simply not worth it.
  • Starlink satellites are already cheap. I don't have any fixed numbers on the exact unit-price, but since they are being manufactured in such high quantities, it is reasonable to assume that they are being mass-produced. As more Starlink satellites are made, individual cost falls and it's possible that per unit manufacturing cost might dip near the price per KG required to launch. The cheaper they get, the less recovery makes sense.

Now, let's look at the price and feasibility of adding a recovery system to Starlink satellites:

  • Reentry: Doable. SpaceX as a company has experience with heat shields and parachutes (Dragon) and it is within their technological capability to equip each satellite with a heat shield and parachute. Regardless though, satellites would likely suffer damage regardless because some parts are fragile (solar panels) and it would probably experience more mechanical stress than during launch.
  • Capture: Difficult, but doable. Using a Fulton-style system where a flying aircraft "hooks" the parachute of a falling object is possible. Rocketlabs has recently even demonstrated so with their first stage, aiming to capture it with a helicopter. Since individual satellites are lighter than first-stage Electrons, a drone (or swarm) could likely capture a satellite in mid air. Drones would probably need to be jet aircraft (or winged in general) due to the large potential area where reentry could occur and the high payload mass necessary. Multirotor-style drones are unsuited for this task because although they have more versatility and can hover, their flight times are more limited, particularly considering that the potential "catch area" would be hundreds of kilometers in diameter.
  • Refurbishment: Doable. I find it unlikely that a 5-y/o mass-produced satellite is valuable enough to repair (eg new solar panels), refurbish, and refuel. Regardless, it is well within the technical capability of SpaceX

Final Verdict:

I don't think that there's any significant technological barrier to recovering satellites. SpaceX has shown us with fairing recovery and with the recovery of Crew Dragon that the fundamental principles are there. That said though, I don't see satellite recovery as being a fiscally wise decision, particularly considering how cheap individual Starlink satellites are--the effort and costs put into recovering and refurbishing a single satellite would likely easily outstrip any potential savings.

Current facts put the price per unit Starlink Satellite price at below \$500,000. More recent estimates still have unit-prices in the six-figures, but rapidly dropping with estimates approaching five-figures. It's not unreasonable to assume that costs will eventually drop to $50,000 per satellite in a couple years, or the price of a nice car. At those prices the prices for, heat shields, drones, parachutes, and recovery operations will easily eclipse the individual unit cost per satellite. Furthermore, even if you do recover a satellite, why use it again? It's likely that the upgrades required to make the recovered satellite usable again would be more expensive than simply building a new one.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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