At least three years ago, I remember talking with some KSAT folks upon Amazon's entry into the ground segment space. While they weren't panicking, they were attempting to figure out what would differentiate themselves from the potentially larger offering from Amazon. I'm not sure they've figured that out yet.
As for the tactical downlink, that needs roasting. Too many companies are making specialized, secure, ultra-slow equipment for soldiers in the field. Now, Ukrainian soldiers are demonstrating what they can do with a Starlink connection, cheap drones, etc. Pretty sure the U.S.'s military network, while it works, can't accommodate such flexibility.
Much to agree with here and looking forward to Part II.
My ha'pworth, tangential thoughts on this evolution, are that the low latency on the fly fusion, #insightsasaservice down, custom processing algorithm up components of a data infrastructure that integrates EO "behind the scenes" is dependent on scalable long-run (sustainable) use cases (and accompanying business models and GTM) and the ability to deliver that across an optimised comms infrastructure.
On the latter had rather taken for granted that the challenges and risks of space based compute at the edge were sufficiently solved to make viable the "squirting" of (relatively) small amounts of insight (from satellite borne DCs - backhaul in a vacuum if you will) down (or across or up) any available comms channel a given?
Not sure why you think space is the worst place to deploy software. Sure, updates may be a challenge, but at least you have a stable hardware platform, no browser incompatibilities, no user interaction, etc.
Very interesting article thank you for sharing. The opinion you share very well describes our work on Space-OS where we develop a lean and secure OS for exactly the kind of applications you mention. I would be glad to have a conversation to learn more
If you have time, can you please explain this statement? "... the larger a constellation becomes in terms of assets on orbit, the more likely it is that at any given moment one or more of them is missing a fleeting (and lucrative) opportunity."
This could be a helpful visual aid - I’m just talking about physical line of sight to a target. As you crank up the number of satellites, you crank up the frequency of revisit for any arbitrary point on Earth: https://www.planet.com/our-constellations/
Thank you Joe. When I read the word "likely" I think probability. Let's say you have a constellation with just one orbit and 3 assets. You will most definitely miss a fleeting event of interest somewhere on the planet. Now let's say you have a constellation with 10 orbits and 100 assets. Should you not have a higher probability of collecting an image from the fleeting event with timely commanding?
Joe, love it! One thing I would add is that the "internet in space" does have direct impact on the ordering piece - because the inter-sat network allows that emergency collection 5 minutes out, your ordering system better be able to plan out that collection FAST in response to that emergency request, while considering all the dynamic stuff that goes with that including which ground station to uplink from, the other emergency orders you have going on, can the satellite do it, is there clouds, what was the satellite planning to do that gets interrupted, etc. etc. Obviously, the only way all this works is if you are automating all that. Hats off if you've solved that, if not... call me ;)
At least three years ago, I remember talking with some KSAT folks upon Amazon's entry into the ground segment space. While they weren't panicking, they were attempting to figure out what would differentiate themselves from the potentially larger offering from Amazon. I'm not sure they've figured that out yet.
As for the tactical downlink, that needs roasting. Too many companies are making specialized, secure, ultra-slow equipment for soldiers in the field. Now, Ukrainian soldiers are demonstrating what they can do with a Starlink connection, cheap drones, etc. Pretty sure the U.S.'s military network, while it works, can't accommodate such flexibility.
Thank you for this funny but accurate description
Much to agree with here and looking forward to Part II.
My ha'pworth, tangential thoughts on this evolution, are that the low latency on the fly fusion, #insightsasaservice down, custom processing algorithm up components of a data infrastructure that integrates EO "behind the scenes" is dependent on scalable long-run (sustainable) use cases (and accompanying business models and GTM) and the ability to deliver that across an optimised comms infrastructure.
On the latter had rather taken for granted that the challenges and risks of space based compute at the edge were sufficiently solved to make viable the "squirting" of (relatively) small amounts of insight (from satellite borne DCs - backhaul in a vacuum if you will) down (or across or up) any available comms channel a given?
Great stuff, thank you for writing!
Not sure why you think space is the worst place to deploy software. Sure, updates may be a challenge, but at least you have a stable hardware platform, no browser incompatibilities, no user interaction, etc.
Haha I can’t tell if this is a joke or not…I’ll talk more about the vagaries of processing in space (vs on the ground) in part II
Very interesting article thank you for sharing. The opinion you share very well describes our work on Space-OS where we develop a lean and secure OS for exactly the kind of applications you mention. I would be glad to have a conversation to learn more
Sounds cool!
If you have time, can you please explain this statement? "... the larger a constellation becomes in terms of assets on orbit, the more likely it is that at any given moment one or more of them is missing a fleeting (and lucrative) opportunity."
Thanks!
This could be a helpful visual aid - I’m just talking about physical line of sight to a target. As you crank up the number of satellites, you crank up the frequency of revisit for any arbitrary point on Earth: https://www.planet.com/our-constellations/
Thank you Joe. When I read the word "likely" I think probability. Let's say you have a constellation with just one orbit and 3 assets. You will most definitely miss a fleeting event of interest somewhere on the planet. Now let's say you have a constellation with 10 orbits and 100 assets. Should you not have a higher probability of collecting an image from the fleeting event with timely commanding?
Possible typo: hydren --> HydRON. However, it's in planning stage and for higher throughput.
Note that currently no system uses optical ground to space links operationally due to atmospheric impairments from air turbulence.
Probably useful to add info re geo relays. There are several of them, RF and optical ones (e.g. edrs-c for sentinel-2 and pleiades).
My bad, yes, I meant Hydron!
Joe, love it! One thing I would add is that the "internet in space" does have direct impact on the ordering piece - because the inter-sat network allows that emergency collection 5 minutes out, your ordering system better be able to plan out that collection FAST in response to that emergency request, while considering all the dynamic stuff that goes with that including which ground station to uplink from, the other emergency orders you have going on, can the satellite do it, is there clouds, what was the satellite planning to do that gets interrupted, etc. etc. Obviously, the only way all this works is if you are automating all that. Hats off if you've solved that, if not... call me ;)