Planet Launches Satellite Constellation to Image the Whole Planet Daily | Hacker News Reader

Planet Launches Satellite Constellation to Image the Whole Planet Daily

planet.com

117 pointspetethomas47 comments

Competitor BlackSky[0] says their objective is "90km^2 in 90 minutes for $90", and they'll be able to do it at 1m resolution instead of Planet's 3-5m resolution.

Not only that, but BlackSky's platform lets you order imagery (both tasking and archive) from many different vendors[1]. So you could set up a recurring job to capture, say, a major port, and the resolution will be good enough to count individual cargo containers (and not just large boats). Then, you could order archived imagery to compare it with.

[0] https://www.blacksky.com/

[1] https://www.blacksky.com/2016/12/14/spaceflight-industries-u...

That's a tasking model. The satellites can't capture everything, so they get a list of specific regions to observe. In other words, you need to know ahead of time what area you want to observe and you'll tell a satellite to point itself towards that area.

Planet's dove constellation is a monitoring model. It's medium resolution (~4m - i.e. you can see a house clearly, but can't reliably see cars), but captures Earth's landmass on a daily basis (modulo cloud cover).

In a nutshell, they're different use cases. A tasking model doesn't help you if you need to know what changed in an arbitrary area a week ago. A monitoring model is less useful for something like a port, where you know you'll want to always image it.

BlackSky will also allow you to task ad-hoc; the plan is to have 60 sats in orbit by 2020 with a ~15 minute revisit to any place on earth.

http://spacenews.com/blacksky-global-says-its-poised-to-cove...

Using the SpyMeSat iOS app ( https://itunes.apple.com/us/app/spymesat/id691290387?mt=8 ) you can also get 1km² or 2km² for 10€/20€ already. They offer access to a lot of different sats, some have a resolution as high as 50cm.

I wonder how many people trade on this information already.

If you Google "hedge fund satellite data", you'll get loads of results, e.g:

> Every five minutes a satellite captures images of China’s biggest cities from space. Thousands of miles away in California, a computer looks at the shadows of the buildings in the images and draws a conclusion: China’s real estate boom is slowing.

> Traders at BlackRock, the money management giant, then use the data to help choose whether to buy or sell the stocks of Chinese developers.

https://www.nytimes.com/2016/11/15/business/dealbook/the-nex...

With this in mind have people been gaming the system?

A building on its own is a liability, its only worth something if it has occupants (or property prices are rising)

China is hard, though, because it's cloudy very often. Having a fleet of satellites in orbit with a ~15 minute revisit rate will be game changing to making sure you find exactly what you're looking for on the day you're looking for it.

Not just for traders, but this is really important for combating fraud, waste, and abuse.

The World Bank, for example, makes lots of overseas investments in places where it might not be convenient to send representatives to check up on the progress of, say, a factory for which they have funded the construction. They'll receive monthly reports about the factory and how many people it's employing, and when they finally go check up on it, it's just a guard house and a pile of rubble.

Having a fleet of satellites that you can use to monitor your investments will definitely make it much harder to get away with stealing money through this type of fraud.

> The World Bank, for example, makes lots of overseas investments in places ...

Perhaps you meant the International Finance Corporation (IFC)?

The World Bank's mandate is primarily to give out loans to developing nations. The IFC is the arm of the World Bank Group that makes investments.

They already do - several years back I was at a supercomputing conference and one of the presenters was processing satellite data, counting cars in parking lots so traders could estimate revenue of certain large retailers. That was 6+ years ago... Imagine combining that with modern ML (not so much for the car counting but the revenue estimates based on that)

Privacy implications are non-tinfoil, rationally-numerous: which homes are abandoned/rarely visited, who's cheating on whom, which homes throw the most parties, parking lot monitoring and so on. Eventually (say a century, if we make it that far), it'll be impossible to go to Antarctica or even wash up on some atoll without literally watched, profiled, surveyed and advertised to.

Actually the search and rescue potential is interesting. Maybe there could be an international symbol anyone stranded can make out of rocks or in the sand or whatever that the eyes in the sky automatically alert the authorities to.

PLBs/ELTs are much more scalable, at present. Someday, maybe AI will "watch" every m^3.

Just don't voice any "treasonous" opinions verbally, in the future, because Palantir Panopticon Next will bring SWATings, tax audits and extra pat-downs to whomever is labelled a foe. Gosh, there might even be a viable, non-tinfoil, future business model for rental of soundproof/TEMPEST meeting rooms in large cities (LOL, I hope not.) Total, real-time monitoring would be a massively-sharp, double-edged sword that has the potential to invade privacy like never before... and I fear people won't fight or even question it.

S.O.S. is the closest thing we have to that symbol. Perhaps we can let CAPTCHA-solving algorithms loose on all that image data...

The bigger issue is that the areas where people are most likely to be stranded are those least frequently surveyed by high-resolution imaging satellites.

And that lots of emergencies will be related to weather, when the terrain will be covered by clouds.

Interesting idea, but it took me about 1.21 seconds to realize the potential for trolling this would have. Plus it would have to have enough entropy to not occur randomly.

I'm imagining a QR code made out of coconut shells.

Very interesting, but how do you dinstinguish from abuse or pranks? It can't be hard to make or those needing recue may not have the resources.

this is an awesome idea

Are there anymore details known, like the resolution of the cameras or what size objects will be visible? With each dove covering 2 million km² per day (it says at the bottom) and capable of 200mbps downlink, one might get an upper bound of what resolution is feasible (with different compression techniques) but I am guessing the limiting factor is the camera.

And does each dove cover the same surface daily, or do they rotate around and get a different patch of the earth every day? Because in the former case, with diffs you could get amazing compression ratios, and I bet you can already get very far with very simple and fast algorithms.

Planet allows free CC-BY-SA access to California region imaginery, so you can see / play / or even develop those algorithms by yourself: https://www.planet.com/products/open-california/

NanoRacks-Planet Labs–Dove collects Earth imagery with a resolution of 3 to 5 meters at a global-scale with a revisit rate currently unprecedented among existing government or commercial satellite systems.

https://www.nasa.gov/mission_pages/station/research/experime... (site has a few example images that look relatively unprocessed as well)

For anyone like me that was initially picturing 88 "large" satellites requiring multiple launch vehicles - this was 88 micro satellites: 10cm x 10cm x 30cm from a single launch vehicle.

On one hand it's fascinating to see the miniaturization of satellite technology, while on the other, one has to think about the amount of "junk" floating around our planet, some of it not much larger than a soda can.

Flock: http://space.skyrocket.de/doc_sdat/flock-1.htm CubeSat: http://space.skyrocket.de/doc_sat/cubesat.htm

Low Earth orbit debris isn't a long-term problem. Atmospheric drag is enough to bring things down before too long. At the ~600km altitude these satellites use, they might last a decade or two before deorbiting naturally. Possibly less, since smaller satellites have proportionally more drag.

Higher altitudes are a much bigger problem, since debris there can stay up for a very long time.

You're correct. In addition, it's important to note that conjunction (collision) risk is proportional to spacecraft cross-sectional area. Thus, even with 100 of these small 10cm x 30cm spacecraft, they still pose less debris risk than a single satellite with a few square meters of solar panels (of which there are plenty already in orbit).

I don't think that's quite right. Say for example a number of small satellites were arranged on the points of a triangular grid pattern, with the points the same distance apart as the diameter of the spacecraft: This would result in a much higher collision chance (for a spacecraft passing through) than if there was a single satellite with the same amount of cross sectional area.

Nice, I never considered the impact of cross-sectional area on the actual collision risk.

Actual footage of the 88 being released in orbit from another thread.

https://news.ycombinator.com/item?id=13653040

Planet is very careful about space junk. One of the founders even did work for NASA, v.v. space junk. All the doves will burn up after use

Well burnup or re-enter

But of course re-enterring with these materials is a burn

Interesting. The data will surely prove to be useful and will also help to show a better understanding of how our planet is changing . Hopefully raw data is released to the public with all the imaging details.

Doves in space

https://twitter.com/dovesinspace

It sound amazing!

Are there any info about the price of those images?

was this among the 104 satellites launched by ISRO (India) yesterday?

Yes.

One step closer to Kessler syndrome.

You have no idea what you're talking about.

Given that now a new kind of space race between private entities seems to be on, and the trend toward micro satellites is continuing, we are going to end up with lots and lots more objects in LEO (and AFAICT these satellites are indeed in LEO), which just increases the potential for cascading failure as Kessler predicted. Or were you assuming that none of these satellites would malfunction?

88 small commanded satellites in a high drag LEO circular orbit contribute to Kessler syndrome in the same way that your production of CO2 contributes to global climate change. Your attitude is just as anti-science as that of people who think Jesus rode a dinosaur.

I appreciate your very constructive comments on my belief system.

But the person you're responding to isn't wrong. These satellites contribute absolutely 0 to space junk.

It was gratuitous, but not everything humans do is bad.

LEO isn't the problem. Anything in the orbit these are in has a lifetime of a few years at most unless they have active propulsion (i.e. not debris or microsatellites).

The problem with space debris is mostly higher orbits where it will remain in orbit indefinitely. In LEO, the atmosphere does a good job of "cleaning up" debris and inactive satellites.

I'm looking at this http://scilogs.spektrum.de/kosmo/gallery/4/liftime_satellite....

I couldn't find a reference for what the orbital height for these satellites are, but I am guessing they are high enough to last decades.

This launch was to 500km, and the Doves are around 150kg/m^2 (4.5kg / 0.03m^2). This puts their lifetime at around 5 years according to that linked plot.

For reference, Planet's earlier launches were to the ISS (400km) and lasted less than a year, which your plot agrees with.

The previous launch for their system has satellites with a perigee of 494 km, so not high at all.

Reply on news.ycombinator.com