I think the word "unmanned" in the title is a bit misleading. I wanted to read the article because I was curious about what I thought would be an AI project. "ATC? How do they pull that off without human operators?"
Such things do exist, at least for general aviation.
This is based on memory from landing (as a passenger) at 29 Palms in California many years ago - hopefully an actual pilot might chip in.
From your map you get the frequency of the tower - when you tune into this there is a looped message giving you some instructions. If you send some click sequence (tap the send button on the radio 3 times or something like that), it plays a weather report.
You have a good look around and if you don't see anyone else landing you broadcast your intentions "intending to land from the north at 29 palms" - then you have a good listen and if no one objects you say your entering final approach, cross your fingers and land!
Sounds like you encountered a SuperAWOS system. When flying, you click the mic three times for weather and runway information, four times for a radio check, and then somewhere around 5-6 clicks you can get the weather/runway information read to you in Chinese. It works great for weather and such, though it doesn't provide any services for separating you from other aircraft.
The SuperAWOS systems are made by a company called Potomac Aviation: https://potomac-aviation.com. David, the head of the company and owner/operator of Potomac Airfield (VKX) is an entertaining guy, his main website is trippy as hell: http://potomac-airfield.com.
Or ASOS. The ASOS is a bit more sophisticated. Both are automated, including the generation of the outgoing message.
ATIS found mainly at airports with a control tower, is recorded by a human, and contains airport information in addition to weather information; construction, hazards, active runways, what approach procedures are being used. They can also be approach and departure specific.
All of these can typically be called by phone as well as heard by radio.
For reference, the overwhelming majority of airports don't use air traffic controllers ("non-towered"). That's because they are way too small and service too few aircraft.
For those airports, the procedure is to dial the unicom frequency that the sectional chart gives for the airport and tell everyone what you're doing.
It's unclear what exactly this control tower actually does for pilots flying into the airport. It's supposed to centralize ATC so they can remotely control air traffic? From the project site, it seems like it's just the cameras and monitoring. No communication otherwise.
Nitpick: Technically it's the common traffic advisory frequency (https://en.wikipedia.org/wiki/Common_traffic_advisory_freque...) rather than unicom. Often they're the same one, but in places where they're split, unicom is for the ground and CTAF for coordinating in the air and for taking off and landing.
The radio communications aren't interesting or novel. The UK physically moved all its (non-tower) controllers hundreds of miles as part of a consolidation exercise.
So yes, the people you're seeing in the photographs are talking to pilots over a radio, they're saying e.g. "Ok, Four Six Two, after that Piper crosses Charlie you're next for runway One Six, let us know if you're ready", because from a hundred miles away they can see where 462 and the Piper are on the airport. "Radio works at a distance" is not a news item. The virtual tower is the news.
My guess is that at this time, they're trying to establish a remote ATC tower can gather and process enough data to be able to do anything for those pilots even in principle. Maybe after that they'll consider communications, because there's no C in ATC without comms...
This is part of a project to provide better aircraft location information in mountainous areas.[1] Radar visibility in mountainous areas is poor. So a system was deployed that pings aircraft transponders, like a full radar, but doesn't have a directional antenna. Instead, there are small transmitters and receivers in multiple locations, and from relative receive times, position can be computed.
So the airport camera system is just the airport component of a larger control area.
Using the receive times to determine a plane's location is one method (multilateration), but it's also becoming common for planes to broadcast their position (ADS-B), determined with a high degree of accuracy using their onboard GPS. This ADS-B transmission is what sites like FlightRadar24 use as their main source, although they're now also able to use MLAT in areas with good base-station coverage.
There has been one in Leesburg, VA for a few years undergoing testing and evaluation. It is the first in North America. It is currently in daily production use.
I wonder what would be the failsafe on these systems? Are there redundant cameras? I would guess there will be dual fiber paths, but other than that? Or is the visual aspect just a relic and can most remote operation be done using radar/telemetry anyway?
The failsafe is pilots talk to each other in the air. There are lots of totally unmanned airports and airports have established procedures to follow if you don't have radio contact.
Not sure about this airport, but in some (most?) places with a tower, there is a very low tech contingency where the airport can operate from a hut on the ground. Everything could be done visually plus a handheld radio and would obviously be slower, but it's there in case the tower had to be evacuated, or similar.
See, for example (there's a picture comment on the question)
> The failsafe is pilots talk to each other in the air. There are lots of totally unmanned airports and airports have established procedures to follow if you don't have radio contact.
This makes sense, of course there will be contingency plans, up to and including shutting down the airport and rerouting flights like your sibling mentioned.
> Not sure about this airport, but in some (most?) places with a tower, there is a very low tech contingency where the airport can operate from a hut on the ground. Everything could be done visually plus a handheld radio and would obviously be slower, but it's there in case the tower had to be evacuated, or similar.
I think this highlights my point, since the airport itself is now "unmanned" in the sense that there is no one there to do the "legacy" way of visual controlling. This means that the analog fallback scenario is not an option anymore, which could be of small importance if it was not used a lot anyway before actually falling further back to other options.
Light signals. All control towers can issue clearances by light signal. Pilots still have to learn about it, although practice opportunities are hard to come by.
Well, what happens when it is too foggy for controllers to see out of a traditional tower? What happens when radar or other systems go down? There will be procedures and fallbacks culminating in closing the airport if they have to.
Instrument Flight Rules (IFR) apply for when it's too foggy: Less than 3 miles visibility and a ceiling of 1000 ft above ground.
When you are on an IFR flight plan and comms fail you stick to your flight plan if still in the clouds and arrive at various point at the time you are expected. Priority during comms failure is landing in good weather if able.
Radar failures happen and there are times when you need to do position reports.
Sounds like a variation on ARTCCs, and TRACON. The difference being piping (hopefully high quality and zoomable) video to the remote "tower" controller, in addition to the usual radar and radio communication.
