Originally Posted by
tallpilot
How many pilots do we need for that 0.1%? It IS coming and most of us will see some percentage of it in our lifetimes. Many will be retired; many more will be past retirement age but still working after getting screwed by multiple downturns and being in the wrong part of the wrong seniority list at the wrong time.
Modern soulless MBAs will spend billions with a B on cap ex to save millions with an M on labor costs. It is the way they roll. Especially if something happens to their shiny buyback scam toy. There is very little Wall Street loves more than screwing labor and automation is just the ticket.
If you are 55+ feel free to poo-poo it all day long. I suspect you are younger than that though. I put it at a 50/50 chance of affecting a 40 year old's career progression and 80/20 for a 30 year old.
Your drone example demonstrates that 1 pilot at each airport is all that is needed for LoS intervention. Add in automated ground vehicles or drone swarms and the number of airports freight haulers need to serve drops quite a bit.
You need one of three things:
1) A generalized AI which can function as a human being in such a manner as to be certifiable. We have no idea how to do that, and we also have some serious reservations about whether we should if could.
2) A dumb automated aircraft which just flies from point A to point B mindlessly and reacts to pre-programmed events from a script. If it gets off script, it just crashes.
3) A sufficiently comms network to ensure an level of QoS and reliability that is currentl inconcievable to allow ground-based pilots to deal with emergencies.
We don't know how to do #1.
We could do #2 right now, and accept that X number of flights will end as smoking holes. You could possibly, with enough redundancy, achieve something in the ballpark of 10^-9 statistical safety, but I think you'd still be one or maybe two decimal places off. The problems are twofold, one is the cost of all the extra redundancy (and operational costs of CANX and diverts when any little redundancy is lost), the other is the psychological aspect... people like control and even if the statistics were in their favor they'd still rather have someone fighting for their lives on the way to the impact site. Many people are fearful of flying, but are perfectly happy to engage in much riskier activities (such as driving to the airport) if their hands are on the wheel.
#3 would be hideously expensive, and would require vast and redundant sat and ground based systems. Sat systems are VERY easy to jam since their power output is minuscule. We have nothing which even remotely resembles the robust comms network required, anyone who uses the mil system would consider it better than nothing but by no means comprehensive and reliable.
Recall that the military lost about half of their UAVs to non-hostile-fire accidents. But financially it made sense to have expendable, relatively cheap drones providing persistent presence over the objective. Also in some cases it was better than risking the life/capture of a human pilot. Airliners are neither cheap nor expendable.
Also I think a lot of people tend to think just because it's possible, that it's imminent. Science fiction writers accurately predicted space travel and nuclear submarines long before they were technically possible... at least 100 years early.
People also forget that even if it's possible, vast systemic changes are vastly expensive... managers will default to what they KNOW will make money next quarter, as opposed to squandering vast sums of shareholder value on risky projects which may reap big benefits for their successors years down the road. Case in point: BCA and the endless 737 upgrades...
Originally Posted by
tallpilot
I'll also point out we purposely program cell towers to ignore airborne targets because it is confusing to the algorithms when the client can 'see' so many towers. Increasing bandwidth and faster processing could change that. Communication reliability is improving.
I've heard that, but it would be problematic for users on high terrain or even skyscrapers... I'm not sure that could be implemented accurately enough to exclude all airplanes but allow all lost or injured mountain climbers, or 911 callers in a high-rise condo. Lots of liability in deliberate exclusion of callers, that's why cell providers didn't auto-block spam callers until the fed gave them legal protections.