HN Explorer

20 comments

[ 4.7 ms ] story [ 51.5 ms ] thread
woliveirajr
> You can build them faster, which means you can put new technology on quicker instead of waiting 10 years for new technology infusion on a government program

I imagine that having one cube launch every year with top equipament (or the best that the budget allows and tha fits the cube) can have better results than one big giant with older technology.

Of course, probably some tech cannot fit a small satellite, but I bet that the vast majority can. And if it can't now, it can in the following 3-5 years, before the big one is released.

Retric
A single cube sat is vastly worse. They are comparing an array of cube sat's which means they also end up very expensive due to short lifespans.
btilly
My big concern is that we still don't have a solution for https://en.wikipedia.org/wiki/Kessler_syndrome. Lots and lots of small satellites brings us closer to that potential catastrophe.

My big hope is that SpaceX's BFR can potentially launch and maneuver cheaply enough to take dead satellites out of orbit in a cost effective manner. Hopefully they manage to launch, and someone is willing to pay to make it happen...

mistrial9
concern ok -- hand-wringing.. not so much.. obstruction and FUD .. you get the idea .. offhand seems like it is a crucial time to get data
gregallan
There are other attempts at disposal of dead satellites that don’t rely on a long-shot like the BFR [1]. Besides, maneuvering a big heavy spacecraft like that to match orbits with a bunch of little satellites would be tremendously wasteful in terms of fuel. A much better use would be to remove larger satellites like defunct communications satellites. These are actually a much bigger danger to orbital space as well. See this event for example [2], which generated over 2000 pieces of debris large enough to track and many more of a smaller size. Small satellites are not nearly as dangerous because when they are defunct their low mass causes them to renter faster.

(FYI I’m a PhD student in the lab that built the satellite in the article.)

[1] https://www.google.com/amp/s/www.theverge.com/platform/amp/2...

[2] https://en.m.wikipedia.org/wiki/2009_satellite_collision

maxxxxx
Most satellites these days have a way to deorbit once they reach the end of their lifetime. I hope this applies to the cubes too.
mirashii
In LEO, that deorbit mechanism is called drag.
maxxxxx
Any ideas how long it takes to deorbit?
AngryData
It depends on the altitude/orbital height. Some are made to come down after only 5 years, I think most are restricted to no more than 25 years and head for a 20 year target. Of course just a small amount of on-board propellant could extend the time further boosting it a bit after a few years of drag.
mirashii
Cubesats at their typical altitudes have deorbit times ranging from 6 months to 20 years, with regulation (in the USA at least) mandating a deorbit within 25 years to receive a license. The fear of Kessler syndrome is much more relevant to launching large geostationary satellites that may break apart than for cubesats.
btilly
That is good to know. Thanks.
Gravityloss
For small satellites and low orbits, drag brings them down.

There is lots of potential for low mass deorbit devices for other satellites. These projects just need funding to iterate them to work.

alliecat
Cubesats are disposable; they're launched in a way that their orbit decays towards the end of their usable life, at which point they re-enter the atmosphere.
interknot
As a software developer, the thing I love about the cubesat concept is the presumably reduced iteration time. The idea of waiting for years only to find out something went wrong once the satellite is in orbit seems absurd (sorry, GOES-17).

That said, isn't the whole concept of cubesats tied to the idea that they're not in a geostationary orbit? There's something to be said for the simplicity of "point at this spot in your section of sky and receive data".

I suppose now is a good time for me to look into how to get data from a cubesat fleet.

geuis
If you're curious about this, check out Planet Lab's. http://planet.com/

They have a fleet of almost 200 cubesats now, in addition to higher resolution satellites from TerraBella. There's an api that you should be able to experiment with.

redavni
Sadly, their time resolution (at least at the free tier) is awful.

Would be quite an interesting challenge to interpolate a large fleet of cubesat's imagery into decent time resolution (the goes satellites regularly do 30 sec time resolution ) over a large geographic area.

edit: The cube sats in the study are only microwave as well, which are compared to the noaa-20 series which arent the correct comparison really.These cubesats would fit in better with the GPM constellation for which global coverage is pretty good, but more coverage would be good. The biggest problem as mentioned is calibration. Accuracy of the dataset over long periods of time is Paramount.

[dead] pokemongoaway
"The bottom line is that this tiny satellite collected data that is as good as the data from a billion-dollar government satellite." -- It is strange that this doesn't trigger anyone else's alarm bells. Did anyone even care to show the exact difference in data between these two sources? Where is the evidence there is a significant figure accuracy difference?