Hacker News Reader

Comment by csteubs | Hacker News Reader

I'm working on a startup trying to make that happen now. There's an immense amount of "free energy" in the form of 120k+ flights around the world every day with that # expected to increase in the future with air taxis, delivery drones, etc. Idea is to hitch a ride on these platforms and crowdsource (cloudsource?) as much aerial image data as possible to create a map that updates every few minutes. Planet Labs and Maxar are great and satellites aren't going anywhere, but we see a pretty big unfilled opportunity in the remote sensing market.

Apologies for the shameless plug--we're on IG @notasatellite if you're interested in seeing some examples :^)

View on HN

Interesting concept, but my gut says it could lead to some malicious use cases.

You're spot on, and that's something I think about daily. Satellite-based incumbents have an average revisit rate (assuming good weather) of roughly 2x/day, so hundreds or thousands of passes/day poses serious privacy and operational risk. We rely on guidance from EthicalGEO and other consumer privacy orgs when drafting standards, but the path to the type of hyper-revisit we're envisioning will need to be tread very cautiously.

How might this work? Could you give example of privacy heuristics you are considering?

Also curious. For example, if I run a shop and I normally have N cars parked out front, but anyone can see with 5 minute accuracy that I always have at least N-2 cars, they could easily figure out when nobody's watching the shop (or house).

Edit: also, nobody can ever sunbathe in peace again.

I would expect that most planes fly on specific "routes", leaving a lot of area uncovered. Do you have any solution for this problem?

You're absolutely right, there are huge swaths of airspace that rarely see traffic. Imaging satellites have an inherent advantage when it comes to total coverage but proportionally, a small % of all images captured in an orbital period are of interest at that point in time. We look at our approach as an opposite, complementary offering that optimizes for revisit in well-established and predictable routes around population centers (and whatever else lies in between takeoff and landing).

Most airports are zoned industrial and located within or on the edge of cities and towns, so we can provide more frequent updates of areas that are of greater value to customers monitoring supply chains, construction/site development, estimating traffic flows, etc. High-revisit is equally important at cruising altitude over much of the country where demand for agricultural/environmental monitoring, crop yield data, etc. is of similar interest.

All this to say coverage is a known limitation, but based on those reasons I think (hope) we'll become the yin to satellite imagery's yang.

Why is privacy intromission so trendy and tempting? I understand this in hard-to-reach areas or for wildlife preservation, but you're effectively adding a real-time aerial Big Brother on everyone's head.

Is technology too widespread? Do we need moral courses, in addition to tech ones?

I understand where you're coming from, but it's much more nuanced than plainly-stated "privacy intromission". For all the good that has come from tools like Google Maps, there's an opportunity to extend that good. For all the nefarious uses, it's an opportunity to pioneer techniques to mitigate those externalities in a way Google and other map data providers have not or will not.

The world is constantly changing and we need better, faster, and cheaper methods of measuring those changes. We can respect (and actively lobby for) privacy while obtaining these results at the same time. Morality and technology should not run parallel to one another, they should both be woven into the products we use every day.

Reply on news.ycombinator.com