How long & challenging training for an IFATC controller isn’t being downplayed at all. The intent of this system is not to replace IFATC. Rather it’s to provide more ATC coverage worldwide in global/live. Right now, that coverage is very small when you think about it. That’s not a ding against IFATC it’s just reality. They do not have the manpower to possibly man all ATC positions across the world at all times of the day.
All controlled airports/fields and areas that don’t have a IFATC Manning the tower, center etc would have Bots. IFATC can configure the Bots to include ATIS messages for different airports
AI is difficult. ATC is difficult enough for a human. The robot would have to recognise takeoff queue order, spacing, pattern order (sequencing), pattern position, taxi conflicts… the list is tremendous. Doing all of this with code is very, very difficult and would require hours of testing. Do you remember the FSX ATC? It was absolutely awful, for this reason exactly.
What happens if someone is approaching at 200 kts and is 1nm from the aircraft in front? What happens if someone takes ages on a takeoff? What happens if someone requests a runway crossing for the wrong runway? What happens if people don’t follow instructions? These are all scenarios a controller can deal with because humans are flexible and can use discretion. A bot cannot and, barring some extraordinary programming, would struggle in these cases.
It would devalue IFATC. What’s the point if there’s a bot wherever you want providing service?
It would reduce traffic even more by spreading out aircraft. Say goodbye to those large ground crews and packed airspaces.
With a bit more API functionality (much easier to implement), there is a possibility an external developer could prototype a system for this. For me, the time it would take to implement even a half working system heavily outweighs the little benefit.
I see what your saying. As I said in another reply, the intent of these bot ATC controllers isn’t to replace or devalue IFATC, rather to supplement IFATC. Like I said before, there’s no way IFATC can possibly control every ATC controlled area airspace in the world at all times of the day: theres not enough manpower.
I’ll be honest, if things stay the way they are now we will continue to have Unicom at most controlled airports & fields across the world in global. I don’t know about you, but having Unicom at most busy IAPs with no approach control, centers etc throughout most of the world is not very realistic nor is it effective for traffic management.
Increasing ATC service worldwide is not an easy problem to solve and I think we need to be creative. If going partially autonomous with ATC Bots isn’t feasible then what other ideas are there?
I’m open to ideas as I’m sure you are. Other than the bots the only other thing I can think of at the moment is that IFATC controllers in centers, etc would have to control aircraft in a much larger area of airspace than they do now. That might cover some coverage gaps in areas that are covered in real life
Intent differs from actual result. If a good AI bot is rolled out, why would I want to go to an airport where IFATC is controlling? People will care less about where IFATC are controlling, controllers won’t bother so much, IFATC is devalued.
Of course it isn’t. But a bot solution would have to be very efficient and robust to have a chance at working. Without serious time investment (time which FDS could be using to implement more aircraft systems, improve lighting- places where it really falls down compared to simulators like FSX), it would make the situation worse.
There is always a steady stream of controllers joining IFATC. The only platform to provide consistent multiplayer, quality ATC is PilotEdge, who charge $35 and only covers the western US, using human voice ATC. No flight sim, including payware has a AI ATC to FDS’s quality standards, and because we already have a controller system in place, to me it would be a wasted investment. Pretty much all the time you could run a A to B between staffed airports, or you can plan a flight to leave or arrive at staffed airports. With the system that we have, I think human ATC is the only real way to get decent service.
I think it would be a fun addition for single player. But it’s almost impossible for multiplayer.
You’d have to programm a tool which is able to handle multiple aircrafts, do sequencing, resequence if one aircraft performs a missed approach, handle aircrafts that won’t follow the sequence, give approach instructions to aircrafts flying at a different speed, …
Programming such a think will most likely take a lot of time and it still will have major flaws since it won’t be able to put all variables into consideration.