2greens1red wrote:
In a dire emergency, requiring immediate landing, how would our robot aircraft liaise with ATC?
Why would an autonomous aircraft require ATC? ATC is by and for humans. Ships, cargo vessels, are becoming autonomous as we speak, no crew at all. Commercial air traffic has not changed since the beginning. It has become more advanced, but the basic structure and principles are the same. Clearly pilotless aircraft will be a huge change, and what eventually sparks it may be a disruptive change in ATC.
Incidentally, the movie “Sully” is all about the “human factor”. In a system that is tailored around humans, I guess all outcome (good or bad) is also dependent on humans doing the right things within what is humanly possible (the essence of the movie). There is no logic saying a robot will function better within such a system than a human. A fully automatic system will be very different, much more efficient, much safer on average, less costly.
LeSving wrote:
Why would an autonomous aircraft require ATC?
It might not, but there would inevitably be a transition period where it will need to work with existing infrastructure.
I know nothing about autonomous ships, but taking a look at the hundred or so cargo ships waiting to unload at Singapore (for example), it’s not easy to see how that could operate without harbour masters and a complex management system (consider poor weather, ‘pirate’ activity etc. – human intervention is essential).
I’d also be interested in hearing how a robot a/c would have managed the Sully incident.
Thread drift: I haven’t seen the Sully film, but fear the worst. It wasn’t a one-man operation – he had a co-pilot and cabin crew…..his adulation isn’t easy to swallow.
There are no autonomous ships, surely.
They are becoming autonomous as we speak…… 
I wonder what will become self-aware first: autonomous AI aircraft or cars?
I always thought it would be a spam filter that would first become self-aware due to the arms race between spammers and anti-spammers – which would of course be a pretty terrible existence.
Peter wrote:
There are no autonomous ships, surel
2greens1red wrote:
Is flying really different from trains/bus transport/autonomous cars?
Cars are difficult because roads are far more complex than air and it’s an environment with loads of unknowns while CAT for the most part has traffic control and even when it fails and collision is imminent pilots are supposed to follow what a box tells them. The issue with aeroplanes is that they can’t just stop. Trains might be the easiest, especially metro/ underground. You have traffic control, you have fixed tracks, they can stop if push comes to shove, but there are some unknown elements, especially if the tracks are above ground (say a tree could fall onto them).
dublinpilot wrote:
What will probably happen is that initially, it will become fully automated from gate to gate, with one pilot still on board in case of malfunction etc.
That’s an issue. Today, if everything is working, single pilot can do it. But when it goes south, even three pilots might have their hands full.
Do you remember Qantas Flight 32? Do you think a computer would have managed that? That it would be able to improvise when, coincidentally, a computer was saying landing was not possible given the damage? There were five pilots in that cockpit and they all had something to do. And they saved over 400 people. What would the programming lead to? Somewhat controlled ditching? Attempt at landing regardless of what the calculations say?
loco wrote:
Same as with current communication loss scenario. Follow the flight plan.
When that plan is for an airport that closed down? Say because of really nasty weather.
So, as I said, it doesn’t exist
what_next wrote:
And in bad weather the landing aircraft will just hit whatever blocks the runway.
How about turning off any landing aids? No lights in bad weather should be a good enough hint.
2greens1red wrote:
How would our robot aircraft have managed? Not just in the air, but on the ground? What’s envisaged for taxying, at an emergency airfield, with one person in the tower, dealing with 14 unexpected arrivals?
The famous closure of Grand Canaria that overwhelmed Tenerife comes to mind. Famous because two 747 that diverted there collided (one attempted to take off in low visibility while another was backtracking and yes, that crash was caused by a human element).
dublinpilot wrote:
The blocked runway is easy to solve. Autonomous cars already have that technology in a much more complicated environment. An aircraft just have to check one straight piece of ‘road’ and go around if it’s blocked.
A car doesn’t have to check a road several kilometers ahead. And why go to cars? Radar is an old friend and plenty of other sensors are used in aerospace applications. It’s IMHO an unnecessarily complicated solution (you would want this if the plane was supposed to be able to select a nice crash-landing site and accurately crash into it or perhaps if you wanted landing capability without any equipment on the ground).
