Pushing Tin Remotely: The Start of Flight Control in the Cloud

In a 1999 movie (Pushing Tin), a flight controller is a passenger on a plane and tells the flight attendant that he needs to speak to the person controlling the plane. The flight attendant tells him the pilot is very busy to which the controller responds, “…you really think the pilot is controlling this plane? That would really scare me.” We wonder what that fictional character would think flying into Loveland Colorado. Their Colorado Remote Tower Project. While there’s still a human flight controller, they aren’t physically located at the airport and rely on remote cameras and radar so the controller can be located elsewhere.

The subject airport is the Northern Colorado Regional Airport and is the state’s busiest airport that has no tower. While the concept — generically known as Remote and Virtual Tower or RVT — dates back to 2002, its adoption is only now starting to pick up steam. An airport in Sweden was the first to go live for normal use in April of 2015, but the Colorado installation is the first approved in the United States. If the official site is a little too dry for you, there’s a CBS report with a video that gives you a quick overview of what’s happening. Or dive in with the demonstration video you can see below.

The technology is straightforward and you have to wonder why it has taken so long. In addition to putting controllers in centralized locations, the system hopes to take advantage of video augmentation, although you could argue you could do that just as well for a controller in place. If you watch the video, you’ll see the centerpiece is a multiscreen virtual window. We wondered if a head-mounted set of VR goggles would have been a better choice.

You can only wonder what will happen first? Complete removal of the human in place of an artificial intelligence like a self-driving car? Or a giant complex of flight controllers in a high rise in a third world country underbidding all the other flight control companies to provide remote services? We figure one of these will happen if not both.

If you want to play at home, [Balint] had a system you’d enjoy, even though the live part is currently down, the project details and videos are still around. We’ve had lots of posts about tracking nearby planes using ADS-B, but one company now listens to all of them.

59 thoughts on “Pushing Tin Remotely: The Start of Flight Control in the Cloud

  1. What is the advantage? Is this technlogie for technologie?
    Having a tower at the airport with people inside is simple and efficient and people don’t stop working because of some technologie failure. Many things could go wrong with all that technologie. Camera, network, power and what else failure.

    1. I guess they can expand their operations without having to build a larger tower. They could have other tower facilities take over in the case of some catastrophe. They can put the tower facility in a more accessible place to make it easier for the air traffic controllers who already work such a stressful high stakes job.

    2. I was a controller a couple of decades ago. I’m torn on this one. There’s likely a cost savings on the human side, but the wide array of possible tech and safety issues is very scary. One thing I’d really be kind of concerned about is the lack of depth perception – judging distances, altitudes and speeds is critical. Shoot… there’s also weather observations, which could be skewed by camera views and monitors, the lack of a 360 degree view (behind the tower is important too!), and how would you handle things like light-gun based control when tower or airplane radios are out? Oy… and now you’ve got cameras, wiring for video and camera control, all the data transfer equipment and wiring, local utility and backup power for everything along the data chain, the display equipment and systems… the list goes on! Oh boy… what about security over all of that equipment and data? I’ve got to stop thinking about the risk factors…it’d keep me up at night, and I put down my mic over 20 years ago!

      Notes on the savings side:
      A physical tower usually has three to four people working – a supervisor, a ‘tower’ controller for planes in the air, a ‘ground’ controller for planes on the ground and a ‘data’ controller who tracks, logs and coordinates. These guys can’t legally work more than 8 hours with at least a 12 hour break, so each tower would need 8 people on staff, 12 if the airport is open around the clock – plus a couple because there would be vacations, sick days and turnover to deal with. Centralizing the staff, you could theoretically have a supervisor cover multiple towers, the data job could either be consolidated or nearly automated (if it hasn’t been already). Ground control could very likely cover multiple airports. The tower controller – I wouldn’t recommend having them cover multiple airports unless they were very, very slow (like rarely have more than two planes in the airspace at a time). The extra person-count that you’d need above the minimums in case of sick time & vacation could be consolidated quite a bit. So there’s a significant potential for savings, but is it really worth it… I don’t know.

      1. “A physical tower usually has three to four people working ”

        Unless it is the middle of the night and pretty much the only traffic is FedEx, DHL, UPS making their one landing and take-off…

        1. Right. I worked the Keesler AFB tower in Biloxi MS, where we’d occasionally have after-hours Hurricane Hunters and Med Evac activity. At those times we were just supposed to have two in the tower for a window of time before/after.

      1. It would need to be more than a power-grid interruption. The two towers I worked in had generators with days of fuel on hand. Both controllers worked on 4 separate radios and each position had access to the other’s radios in case of hardware failure (like lightning dancing on the panel – it does happen). In the event of all of those radios failing, we had battery powered VHF and UHF radios. In the event of all of that failing, we had a ‘light gun’ for basic control communication with pilots if our (or their) radios were completely dead.

        1. Yep, double redundancy…
          The FAA 4 megapixel (ISIS) display introduced in the early 1990’s could swivel left or right to allow one controller to sit between 3 displays (if needed).
          To get a bit political, I think the only good thing Federico Pen~a did as Secretary of Transportation, was end the “FAA Modernization Program” which at the time was $1Bn (1990’s dollars) project that was already $1Bn over budget with no end in sight.

  2. No, head-mounted display *would not* have been a better choice. At least none of the current ones.

    The image clarity is much worse (reading text in an HMD is very difficult!) and the fatigue from wearing these is much higher than from using conventional screens, both because of the eye strain but also the weight and heat over time. Few people can handle wearing an HMD for more than an hour without break while these ATC controllers are on duty for several hours at a time. And doing a much more stressful and complex task than shooting some VR zombies where missing a shot has no consequences. If the controller misses a new blip or alert on their screen, real people may die. So adding anything distracting or introducing more fatigue to an already difficult job is a terrible idea.

    Also keep in mind that the ATC controllers are not only looking around! They need to actually use keyboards, radios, other equipment they may have around, deal with flight progress strips (these are often actual paper strips!), manuals and paperwork on the desk, etc. Head mounted display would impede such work. Even a see-though display like a Hololens is a royal distracting pain when trying to operate a PC, with a non-see through HMD (such as Rift or Vive) you can’t see anything around you. When two planes are on a collision course the controller doesn’t have time to hunt for a button to enable a laggy camera to let him see where his microphone or keyboard is and then try to find them by touch.

  3. The tower at our local airport was abandoned years ago. Currently air traffic control is handled by controllers 90 miles away. As I understand it, they have radar video and radio feeds straight to a console in their tower.

  4. a few things that come to mind ..

    No depth perception .. everything is displayed on 2D monitors. (perhaps this is not an issue since most objects that need to be visually tracked are at some distance away ?)

    How is the performance of the video system in heavy weather with low visibility, does it outperform just looking out of the window ?

    more automation also seems to imply longer downtimes when there are IT problems..

    1. There’s a ground based radar(ASDE) that they use to watch where planes are on the tarmac after they are handed off by the carriers. Funny story, when the Philly airport had that incursion of the SUV on the runway, it was the radar and intuition of the controller, not visuals, that made the controller wave off the landing flight and save tons of people’s lives. These guys are special people. Beasts of executive functioning and working memory. Anything like this is thoroughly tested to make sure it doesn’t increase the workload of a controller in massive studies. We used to study the impact of just a single UI element change like a pop up to send a command to a pilot.

  5. At most major airports the tower itself is normally only used for ground control. The good visibility is useful in case anybody is where they shouldn’t be for whatever reason.

    Air traffic control information has been shared between multiple operators for decades. If you think about it with planes taking off and/or landing every 60 to 90 seconds it isn’t really possible for one air traffic controller to monitor everything. So the radar (actually normally transponder data, supported by radar) is shared to the appropriate controllers who can normally filter the view to their area of operations. This might be the routing into one of the feeder points, or controlling flight lanes around high traffic zones, take offs, landings, etc etc.

    Extending this and moving it to a physically different building is not difficult, and is the normal course of events.

    Boards etc do fail, and it is normal for facilities to have a backup boards, backup ATC rooms and even entire backup facilities that can be spun up quickly.

    If it has been moved to the internet that is just bat shit crazy idiocy of the first order.

  6. Same as remote security(?) porters(?), don´t know the correct word, in buildings. It is all nice while it is only one building being managed, but when they start to make the same group of people manage many buildings just to improve their cash flow, then the service quality falls very fast. How about the remote flight controller telling a plane to land in a landing strip that doesn´t existe in the airport he is managing now, but existed in the last one he managed a couple of minutes ago ? Not nice.

    1. A controller has to be validated on each airspace/aerodrome, and the concept of “chopping & changing” within any densely used airspace is not ever going to happen – there are both national and international regulations which have to be followed – I tell you this as someone who works in the business.

  7. About time..
    A few years ago, flying into “Watson Lake” airport in Yukon Territory Canada, no tower, no radio, no functioning beacons, nada.
    The trip north through the Watson Lake trench was VFR only with a Xerox copy of the lumber company built dirt strips.
    Again, no radio, no towers, no fuel, no food, no nada.
    In Canada, a “Long Gun” is required in the plane so you can eat if you live through the crash.
    But, it was a trip of a lifetime.

    1. Flying out of Glendive, Montana on Northwest Orient back in the late 1980’s.
      There was one woman at the airport, she was the ticket office, baggage handler, ground control, tower and ramp operations.

  8. For small airports it’s maybe a good thing, but overall it’s idiotic. There is no way this thing is safe against cyber attacks. Why not outsource nuclear missile control to the cloud as well?

    1. What is safe? Shall we get rid of all the computers in use in air traffic control? Let’s only use binoculars and semaphore?
      Realistically these systems will use redundant private networks with multiple carriers, multiple layers of uninterruptible power supplies and generators, and redundant ranging and radio equipment.
      That’s not to say it’s perfectly safe, but it’s at least at no more significant risk than existing systems.

  9. A little backstory on one of the reasons why KFNL (Northern Colorado Regional Airport airport, Loveland CO) is trying this. Allegient Airlines used to have passenger service into FNL, but the pilots of the big birds regularly flew against the prevailing flight pattern to save a few minutes. Scared the jeezers out of the small craft pilots, and they complained. Allegient pulled out their service claiming “no tower=no safety”

    Probably an ideal site for a remote tower experiment. 1 runway (really 2, but the crosswind runway is short and rarely used). Lots of executive jets flying in all the time in addition to the flight schools in the area. There’s a major hospital right next door with rotary-wing operations. Denver Center (the regional ARTCC) is just down the road in Longmont (where the remote controllers could potentially sit). The field is easy to navigate to (lots of pilotage landmarks, VORs, and ADBs in the area)
    Basically, there’s a fair amount of air traffic in the area that could benefit from any type of tower, and having some kind of controlled air space would be an advantage for getting a passenger carrier back on the field again.

  10. The FAA has for years been wanting to take local ATC away from Boise, ID and relocate it to Salt Lake City. Their argument is that it will “save money”. At most two jobs would be eliminated and the rest relocated. Then there’s the expense of new hardware and secure + redundant communications links over 350 miles.
    No way is that going to reduce any cost.

    Pilots like the zero delay ability to ask the tower what the immediate weather and ground conditions are, when a controller can simply look out a window.

    1. “when a controller can simply look out a window”
      If you are talking about someone in the top of the tower, yes.
      Otherwise, controllers are in a darkened, windowless room watching displays.

  11. Actually this article is incorrect. First off our Remote Tower and controllers will be physically on the airport, secondly there will be a team of 4 air traffic controllers to man it, NOT just one. Once the FAA has certified it then the attempt will be made to pipe in the data from another airport so that more than one airport may be controlled when the system is not being utilized for our airport. The advantage is a HUGE cost savings over construction of a brick and mortar tower. Additionally we anticipate Remote Towers becoming part of the Contract Tower program, another huge cost savings as that program would also cover the cost of personnel just as they are with conventional towers. Another similarity to conventional towers, ours will have operational hours that will switch to VFR when out of service. You can read all the FACTS at http://www.ColoradoRemoteTower.com

    1. This is some good info, thanks! Just out of curiosity, will the remote controllers have to get rated fully for every tower they cover? It’s been over 20 yrs, but I remember having to memorize frequencies, obstacles & visual reference points, approach & landing patterns, noise abatement zones, et cetera et cetera…

      1. Hi Tim,

        Our controllers will have the same certifications as they traditionally do, there has been discussion on whether our Remote Tower will become a center for certification for controllers for Remote Tower operations but the intent of the creators is to layout our control room exactly like a conventional tower. We will be having our first tour of the control room open to the public at our next Commission meeting, which is October 18, https://www.flynoco.com/about/public-meetings

  12. OMG, the airport tower is there NOT for the ATC (Air-Traffic Controllers); it’s there for the GROUND Traffic Controllers. The ATC is usually not at the airport AT ALL; for example here at PRG the ATC for the whole CZ airspace is actually some 10 miles away. Additionally, they are usually not even the same subject; one company runs the airport with its ground crew, another runs the GTC, and yet another runs the regional ATC.

    1. Not exactly…the tower has a ground controller AND a local controller who handles air traffic inside 5 miles. The radar rooms can be just about anywhere in the vicinity handling the ‘terminal airspace’ (arrivals/departures/overflights). Beyond that are the area control centers (we just called them ‘centers’) that handle broad stretches of airspace between terminals.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.