6 hours ago, tup61 said:

Do you perhaps mainly fly in the USA? What default ATC does NOT do is use the correct transition altitudes: it uses 18.000 ft all over the world. It also uses inches all over the world while hPA is used almost everywhere outside of the US. Default ATC also doesn't use FAA and ICAO terminology (where appropriate). The fact that you say you need to ignore things exactly shows what's wrong with default ATC. And calling out descent points every time while flying a STAR...? Is that realistic? And what about the limited amount of devault voices, switching frequencies way too often, disregarding terrain...? Obviously default ATC also has no speech recognition.

If default ATC works for you then that's great but for most people it doesn't work too well. 😉

Yes, 70% US.  Yes, those are all problems outside of the US.  Calling out descent point depends on the airspace. Denver, yes,  Phoenix, yes they would call out descents and speeds, not necissarily based on STARs but usually.  

Small airports like Boise Idaho,  Tucson, Arizona no.  but again... it all depends on the situation in the real world and the wild card is AI aircraft behavior.

I'm not saying default ATC works well but it works as well as it ever has for a fixed and infelxible system for free.  I'm saying there is nothing I have seen that is groundbreaking in this product...yet.  The accents and voices are cool (hopefully it will do a southern accent in Atlanta or Dallas).  I'm not that into talking to it, if it's not a real human, but it is a cool feature. I let the AI co-pilot control the radios so I'm not sure I'd use it.

Like I said, in a perfect world, every area of the US or Europe would have VATSIM coverage at specific and reliable times.  I would certianly pay for that. Unfortuantely, unless your flying from LAX to LAS or SFO chances are slim you''ll have full vatsim coverage.

I have just been burned by so many 3rd party ATC programs over the last 15 years that I'm immedeatly skeptical when I see a new one.

 

30 minutes ago, BrammyH said:

Did it actually perform the go around?

The plane didn't land but I didn't watch it to see what it did this morning.  However, I see go-arounds being acted on regularly and I also use FSLTL.  I'll take a screenshot next time I see this.

Edited July 11, 20232 yr by Noel

4 hours ago, BrammyH said:

Did it actually perform the go around? I use FSLTL to inject the traffic. It did correctly give a go around for an aircraft while I was lined up and waiting. That plane just went LOL No and landed anyway.

The air traffic control needs to actually control. If I am expecting vectors (usually noted in the MCDU flight plan) it should be vectoring me. When I tell it to give me vectors, I think it just says "go direct to <waypoint name> if an ATC program can't control the AI craft, it at least needs to be aware of them, and their flight plan, and steer me around them. Too often it tells me "Line up and wait. Traffic is airbus on final" and then tells the airbus to land. It's not an FSLTL issue I think, since after it injects it the AI just takes them over i think.

I posted this in the other B-ATC thread so I'll post it here too, because it answers your question.

In the Discord, the lead Developer of B-ATC answered why his program doesn't control the AI, and this was his answer:

Quote

I do have some ideas for how to do actual traffic controlling.

Right now, we need to pick our battles. I need to get the core functionality working. People have expressed they want IFR and VFR as inclusions in the base package, so that needs to come first. We will handle traffic as a layer that sits above BATC. If there is traffic on the runway, we won't get you takeoff clearance, but we also can't space them correctly. That will be 1.0 of BATC. Then, after that's working to our desired degree of quality, we will dive into the world of traffic controlling.

Which, if you think about it is almost an entire other program - how to route planes around an airport is a huge undertaking, never mind having them communicate properly. So that's why I'm thinking we split it up into two parts.

1.0 = IFR and VFR flights with awesome quality phraseology and procedures with a immersive level of traffic interactions, but no direct control.

2.0 = Full traffic control, the entire experience top to bottom.

It might not even make sense to start working on direct traffic control right now - bending over backwards to work with an SDK made out of duct tape and swiss cheese. Then 2024 drops and they expose some amazing new layer that allows precision control, what a waste that would have been for us to focus on the wrong thing first. So it makes sense to me to punt this task until we know more about 2024.

So AI-Control is coming. Either Asobo will do it next year for MSFS2024, or B-ATC will do it themselves once 1.0 is released and Asobo shows no intent of providing AI control. In either case, it'll be a part of the program eventually. Devs always make promises, so take his claim with a grain of salt.

That makes sense.

I am hoping that maybe in MSFS2024 they will have an API for AI control.