[csteubs]

There's still money on the table for airline operators where routes are concerned. It sounds a bit fringe, but most of these flights take off and land at airports in large cities. International airports are often zoned around industrial centers that are typically set off away from the city core. Coupled with the high frequency of inbound and outbound traffic depending on the airport, commercial flights are excellent candidates for capturing high resolution aerial imagery in near-real time. In many cases, those images can be transmitted back down while the plane is in the air thanks to in-flight internet.

There have been past attempts to harness the passive residual capabilities of air travel for photography, but none have stuck based on the difficulty of scaling such a system. If successful, crowdsourced aerial photography for cartography could add a layer of temporal resolution we just can't achieve with fewer than 500 imaging satellites currently in orbit, and for 1/10,000th the cost of satellite-based imaging. Add on concerns about LEO crowding and the thousands of use cases near-real-time aerial imaging could provide at a fraction of the cost, it becomes an interesting problem to solve.

There are plenty of barriers to this kind of scheme, regulation in particular, that could keep airlines from participating; the first major carrier who does, however, may find themselves with a new multi-billion dollar revenue stream.

(Disclosure: Founder @ https://notasatellite.com)

[lastofthemojito]

That's a neat idea, and something I'd never thought of before.

I'd imagine the customers here would be looking for a way to keep an eye on competitors? Ford would like daily overhead shots of GM's factories and parking lots, McDonalds would like to check on the length of average drive-through queue at competitors, etc? All in near-real time, without waiting for quarterly earnings.

On the other hand, weather might render this too unreliable for the needs of such customers - "how busy have the Bay Area Starbucks been since they've released their Pumpkin Spice Latte?" "Uhh, not sure boss, it's the 30th cloudy day in a row, we can't see anything".

[csteubs]

Exactly--one such example is Toyota's export lots sitting in the western approach path at PDX, and Liberty International sitting right next to the Port of Newark. We have a ton of examples on social media if you care to see more.

I've been trying not to get too in the weeds with use cases right now as we're building because they really are endless, but supply chain intelligence and insurance are massive categories based on those who have expressed interest.

Weather is certainly an issue, but one that is shared with optical imaging satellites. SAR definitely helps with clouds and night shots, but with 100k flights per day the odds of getting at least one usable shot from a plane is far greater than a satellite due to revisit rates in orbit not matching weather changes in the troposphere.

[nradov]

Most commercial flights operate in a small number of fairly narrow corridors. Is there really a billion dollar market for aerial imagery of those flight corridors? Seems unlikely.

[csteubs]

Maybe--time will tell. We run revisit estimates with FlightRadar so I can say that there's a good amount of deviation in routes generally, but I also think narrow corridors are the mechanisms that makes rapid revisit even more powerful. When you can achieve minute-level temporal resolution for heavily populated areas using smartphones (to start), the cost is more or less negligible compared to satellite imaging.

I know of one market-maker HQ'd in Manhattan that currently pays $16 million/year for two satellite images per day of the oil storage tanks in Cushing, OK. Analysts determine tank volume by looking at the shadows cast by the floating lids on the tanks and use that data to set oil futures prices. There are roughly 1,500 flights over the same area every day. We're very much a hypothesis-become-startup at the moment but I'm optimistic!

[flavius29663]

This is a fascinating idea, congrats. I see major issues with the existing planes: they do not seem designed to be retrofitted with down facing cameras.

Also, near real time HiRes images means a lot of bandwidth that you don't usually have in the air.

[csteubs]

Thank you! You're correct about (most) existing planes--from what I understand, there was a "gentleman's agreement" between cold war powers that kept manufacturers from pursuing those designs. Since 2000, many models have added onboard cameras on the belly and tail for pilot awareness and in-flight entertainment systems.

Our system is 100% internal which allows us to sidestep massive expense that comes with tunnel testing, retrofit, maintenance, etc. Customers using our Beta receive free in-flight wifi for recording out of their window seat for short increments during their flight. We cache the previous image and compare with the new image on the device itself, transmitting only the raw diff which dramatically lowers our bandwidth needs. All orthorectification and advance photogrammetry occurs on-ground, but we outsource a ton of that processing to the edge device.

We're also designing a physical hardware camera array that can be fixed to the window and is about the size of a headphone case. If you're interested in examples, our social media accounts are full of them: https://www.instagram.com/notasatellite/