Dispatch Flight Plan Automation

Hello all,

New to the forum. Just got a job as a flight follower and am just learning about the world of dispatch. With this hopefully being my career, I have a question about the long term viability of the job.

I went to school with a person who got a job at Allegiant a while ago. They have told me the airline is trying to automate the job to as great extent as possible (I would like to hope the intentions are good and they aren’t trying to eliminate jobs but who knows.) I know something like this would probably happen in the future with AI, but not this soon.

What are your airlines doing in this respect? Is this something to be worried about in the medium-long term? Do you think the FAA will ever let this fly?
 

4EngineETOPS

Well-Known Member
Considering the level of IT incompetence and bureaucracy at many companies (including airlines), I don't think fully-automated dispatch is anywhere on the horizon. Dispatch software is definitely continuously improving, but technological progress, while rapid in a big picture sense, is not as swift as people might think. Aviation tends to drag behind the Silicon Valley cutting edge and AI is still something that is continuing to evolve and develop. Even when we get to a place where AI is able to support fully-automated dispatching up to the standards of humans, it will take years of software development, testing, and regulatory approval. I don't see the FAA removing the operational control requirement anytime soon. NextGen won't be fully implemented until at least 2025, and we're still a few years away from eliminating paper strips in ATC facilities.

This is not to say that technology moves slowly. We've gone from the introduction of dial-up internet on CRT screens to near-5G and speeds on incredibly powerful handheld devices in about 30 years. AI has come a long way in recent years to reliably enable near-autonomous vehicles. I just think there's a big gap between emerging capabilities and taking the leap to full automation. Fully autonomous aircraft would need to take the giant leap from having the technology available to proving that they are fail-safe to the point that 100% of QRH non-normal situations can be resolved without any intervention. For the most part, humans can pilot an aircraft as long as the flight surfaces are controllable (i.e. minimum hydraulics), even with numerous other system faults. Can AI do the same reliably without a compromise in safety? The challenge isn't getting from point A to point B routinely. It's identifying, solving for, fail-safing (?), and exhaustively testing all conceivable non-normal situations. The same applies to completely automating dispatch. It's one thing for software to be able to select a route and fuel policy based on preset rules, and avoid areas of certain EDRs and Doppler radar return thresholds. It's another entirely to be able to consider and incorporate numerous big picture items, effectively respond to non-normals, and pass relevant information along to pilots at the right time. Do I think that aviation will be entirely automated in the future at some point? Yes, like most industries, machine learning and AI will eventually reach the right point for full automation capability, and the issue of interfacing/inter-system communication will be solved. I just think it's a while down the road, to the point that I'll hopefully be retired by the time it happens (and I'm in my 20s).
 

QXDX

Well-Known Member
Hello all,

New to the forum. Just got a job as a flight follower and am just learning about the world of dispatch. With this hopefully being my career, I have a question about the long term viability of the job.

I went to school with a person who got a job at Allegiant a while ago. They have told me the airline is trying to automate the job to as great extent as possible (I would like to hope the intentions are good and they aren’t trying to eliminate jobs but who knows.) I know something like this would probably happen in the future with AI, but not this soon.

What are your airlines doing in this respect? Is this something to be worried about in the medium-long term? Do you think the FAA will ever let this fly?
So who says eliminating jobs is a bad thing? Would anyone argue that society is worse off because there are fewer buggy whip makers and steam locomotive manufacturers? Creative destruction is a feature of healthy economies, and is what improves society's standard of living.

However, to answer your question, I think the dispatcher's role and responsibilities will change, but not go away. The need for someone to sit at a computer terminal and punch out a flight plan may disappear, but not the need for someone to exercise operational control. That may mean the need for fewer dispatchers, but not no dispatchers. Time will tell.
 

bbmikej

Well-Known Member
So who says eliminating jobs is a bad thing? Would anyone argue that society is worse off because there are fewer buggy whip makers and steam locomotive manufacturers? Creative destruction is a feature of healthy economies, and is what improves society's standard of living.

However, to answer your question, I think the dispatcher's role and responsibilities will change, but not go away. The need for someone to sit at a computer terminal and punch out a flight plan may disappear, but not the need for someone to exercise operational control. That may mean the need for fewer dispatchers, but not no dispatchers. Time will tell.
There are airlines in Europe that are already at a similar point as you describe. I was at C5 a few years ago as they were changing dispatch software and the one they brought in was able to plan fights. The big issue is that it was designed in Europe and did not understand the nuances of American airspace and weather. I don't remember which low cost airline it was, but they essentially have a computer plan the flights automatically and then release them with just 1 person overseeing the operation; making sure the computer didn't mess and handling the unplanned situations.
 

manniax

Well-met in the Ka-tet
There are airlines in Europe that are already at a similar point as you describe. I was at C5 a few years ago as they were changing dispatch software and the one they brought in was able to plan fights. The big issue is that it was designed in Europe and did not understand the nuances of American airspace and weather. I don't remember which low cost airline it was, but they essentially have a computer plan the flights automatically and then release them with just 1 person overseeing the operation; making sure the computer didn't mess and handling the unplanned situations.
This actually isn't a new concept. The IT part of it might be, but I worked at a 1900 operator back in the late 1990's before the 121 transition that did the same thing with something called "center stored flight plans." Basically the same flight plan for each flight was filed automatically every day and all you did was flight follow. However for something like that to happen today in the US, they would have to rewrite the "joint operational control" requirement of 121, and I don't see that happening any time soon. Also, as you mentioned, European weather is a lot different - it's far enough north that in the summertime they don't tend to get long lines of thunderstorms, etc. and their dynamic rerouting requirements are a lot lower as a result.
 
Last edited:

Delta Echo

Well-Known Member
I think fully automated aircraft and dispatching will be something not even remotely allowed until maybe the late 21st century. By then most of us will have flown west or have been retired for several moons.
 
Top