I worked on this at Facebook very briefly. I believe the project was called Aquila internally.
I worked with a famous engineer who joined Facebook from a well known british aircraft manufacturer. He was an aerodynamics expert and was trying to solve the problems of flutter and other dynamic instability in the extremely large, high aspect ratio, lightweight composite wing.
I wasn't officially told this, but I believe the reason the project failed was two fold:
1. Traditional satellites became cheap enough that high altitude pseudo satellites weren't worth it anymore.
2. Facebook free basics (internet.org) achieved enough success partnering with traditional mobile network operators that it wasn't needed.
I have some pretty crazy anecdotes to share about the project if anyone is interested?
One last thing that is a little bit sadder, so I thought I'd post in a separate thread.
I grew to really like the engineer that invited me to the project. He was a well known expert in the field with a wiki page, books, academic eminence and everything, and he offered to mentor me etc.
I travelled with him a few times from FB's London office to the prototyping site and he was always very generous with hotels, travel etc.
Eventually someone pulled me aside and mentioned that he had "alterior motives" and had repeated the same pattern with previous young men. I was honestly so saddened and betrayed, and afterwards couldn't believe I had been so nieve as to be made to feel special by him.
That's terrible, thanks for sharing. I thought I was pretty cynical, but I didn't see that coming at all -- good reminder to keep an open mind and always maintain boundaries, just in case. Also the implicit ending is sobering: he likely continues to pursue such asymmetrical relationships to this day, because "cancelling" high-level corporate workers is a lot harder than #metoo makes it seem
In the UK, we had a smaller scale model of the aircraft that we used for testing. The team that built the scale model was having trouble configuring the autopilot and I volinteered to help
They were using an off the shelf 'drone' autopilot based on an open source project PX4, which I had some experience with.
I was asked to visit the location with the prototypes and spent the weekend trying to get their RF, magnetometer and GPS calibrated and working with their carbon composite prototype in the way.
I was expecting that they would already have an expert working on this, but instead I found some very senior hyper specialised engineers that were experts on helicopter rotor design for example, but could literally not read the instructions on an FTDI based usb to serial cable and connect their laptops to the autopilot.
It was a classic situation where you connect the projector in english class and the teacher thinks you're the next Donald Knuth.
I am possibly wrong about small details, and would appreciate corrections.
So when I joined the project, the physical aircraft was based in the USA, but the software team for the autopilot was based in the UK.
For some reason - I think it was the very high altitude capability - the USA considered the autopilot to be munitions.
When I first joined the project, there was this bizzare situation where it turned out to be simpler and easier to occasionally ship the entire aircraft across the ocean, than to send software updates over the internet.
I'm not sure how long this lasted but I was assured it was temporary.
Hey dang, this is the second time I had to vouch for this user's comment. I don't know how your system is set up, but it seems odd to me these are immediately showing up as dead.
Re' sending an entire aircraft over the ocean vs. sending software over the internet: you'd be surprised, but it's still often the case.
I worked for a storage company that made a long-distance replication product (i.e. a program that allows one to store their data both locally and far away in a reasonable time-frame). We were contacted at some point by a movie production company which had this problem of moving the "raw" footage from the place it was filmed to the place it was supposed to be edited, and because it would take ages over the Internet, they'd use an airplane to fly it between the sites.
In general, it's a known phenomenon in the storage industry and as technology advances gets rebranded from "Boeing loaded with CD-ROMs", to loaded with hard-drives, to loaded with USB sticks etc. In other words: Internet is slow, if you want to move a lot of data, as in it has very low bandwidth.
I don't think the size of the update was the issue here.
I guess it's something like the US Export control systems for arms, but for imports? I'd love to know the details of what legislation it was under - possible it was on the UK side.
In my home country, there was a stunt where a person used a carrier pidgeon with a usb drive to race a local ISP transferring a file.
It got me thinking about strange units. Like how for example a car's mileage can be thought of as a surface area (distance per volume), or in this case how for every pidgeon speed, drive capacity and length there is an equivalent bandwidth capacity * (distance / (distance / time)).
> the USA considered the autopilot to be munitions.
I could be wrong but it might be related to the GPS. Most off the shelf GPS units have a height and speed limit to stop them being used in ballistic missiles.
Yes, FB were trying something similar to this with Aquila. We actually used to work with a few people on that project, it sounds like the all composite flying-wing design was a bit of a challenge for them. I also understand they weren’t providing internet direct-to-device, just as backhaul for cell towers, which proved to be a difficult business case. In the end they partnered with Airbus on some projects, but I don’t know what came of that.
I worked with a famous engineer who joined Facebook from a well known british aircraft manufacturer. He was an aerodynamics expert and was trying to solve the problems of flutter and other dynamic instability in the extremely large, high aspect ratio, lightweight composite wing.
I wasn't officially told this, but I believe the reason the project failed was two fold:
1. Traditional satellites became cheap enough that high altitude pseudo satellites weren't worth it anymore. 2. Facebook free basics (internet.org) achieved enough success partnering with traditional mobile network operators that it wasn't needed.
I have some pretty crazy anecdotes to share about the project if anyone is interested?