Yeah, plenty of lessons learned so far, but I don't really see a lot of large commercial unmanned systems using regular airports for a while

There will have to be a lot more failsafes and better training before then.

One of the previous systems I operated had a crew of 4 (three in the ground station). It worked well. Now, as a cost cutting measure, the current system I'm on is single operator. Depending on mission complexity, it's often too much for one person and requires assistance from someone else.

I would consider two crew members to be absolute minimum even for a simpler system, so... going unmanned doesn't save too much on crew costs.

Smaller birds in remote areas? 100%. Something Caravan-sized delivering stuff out in the boonies would certainly be feasible imho, especially when we start building dedicated unmanned cargo birds that don't have to be concerned with seats, oxygen, instrument panels and all the other stuff required for human pilots.

No, you see, that's different......

An airplane built in the 1970’s with a single axis autopilot (HDG/NAV), altitude hold, and Garmin 430 (1990’s tech) can basically be flown this way.  The pilot hand flies takeoff and initial climb, pattern entry/approach, landing, and manages altitude changes.  It will follow set waypoints while en route.

Now make it a two-axis autopilot and you can program your altitude on the ground, change altitude and manage vertical speed with a button, and it will fly basic approaches.  The pilot still manages the engine and throttle.

Newer autopilots can fly the full approaches, have speed and roll protections, and auto throttles.

Ironically the more sophisticated of the two autopilots that I fly (00’s tech) requires more babysitting as it can fly the airplane into a stall.  The older one will just disengage before it does anything crazy.  In any event, the pilot should always be actively monitoring the systems (the babysitting part).

I flew several hours yesterday and probably hand flew the airplane less than 10 minutes. And I like hand flying, but we were going point A to point B.

This is the reason I think it will never happen. Not a pilot and not in aviation with no background and I don't pretend to know anything about it, but I work in cyber security and your comment on the security of pilotless aircraft is spot on. It's unlikely that it will remain secure enough to operate without risk of compromise

Even nuclear plants are vulnerable despite the efforts to secure them. I can't imagine aircraft are going to be more secure

See How to Fly the Pattern Without Making an #@$ of Yourself.

The amount of touch labor still inherent with an aircraft with A/THR and dual CATIIIb capable autopilots is still significant in a place like the terminal environment.

Its a wicked problem, where you've necessarily have to have open network because you're flying over places like Russia or China, but be secure enough to keep even script kiddie level threats at bay.

This is separate from just brute force jamming or other kinds of EA, especially in the terminal environment. The amount of jamming required to deny/degrade something like the GPS signal (which is a critical element not just for navigation, but timing across networks) is really really low. Like a millionth of a watt low.

Yes I can imagine. And to be clear I have zero experience with those systems. My point was that even the more basic systems require significant pilot input and management.  The more sophisticated the AP, the more attention it needs is my guess.

I do have experience with “now let’s see if it captures” as two pilots sit there looking at it during an approach.

The idea of an autopilot malfunction becoming a potential loss of aircraft problem is kind of scary. There have been more than a few times in my career where I just didn't like how the A/P was doing stuff like capturing, or was porpoising on the approach, and I just said "Autopilot's coming off" and hand flew a basic visual circuit.

I don’t focus on the legislative side because there is no need to.

When the airlines, insurance agents and engineers agree its time. They will buy off the legislators,

Anytime someone says “equipment requirements”  it’s a giveaway that they have just enough knowledge to think they know what’s required.

In truth, for just a few hundred thousand dollars I can equip any modern aircraft with the capability to fly totally unmanned.  They leave the factory, as you mentioned, with 95% of the capability required.  The recent ADSB mandate went a long way to making this possible.

The “next gen” ATC system will be the single largest player to moving this forward.  


That’s cute and all, but we gave that to them.

One airplane I fly has an altitude hold thumb button on the yoke.  Some have mistakenly hit it on base/final and engaged the alt hold.

My rule is that the alt hold is turned off with the panel switch anytime it is not being used.

Nothing like fighting with the AP in a critical phase of flight.

Regulation and legislation will be a large hurtle if not the largest, doesn't matter how much money the airlines spend (and by the way the pax guys won't have much spare cash for some time).  Regulation will tie into requirements and testing.  I am sure some dude who lives in his mom's basement could have easily made an and comparison in the FCC code for the max.  Though all the regulation, testing, and requirements that came after are what drug it out and costs money.  You think Boeing and all the other US airlines weren't trying to call in favors?  Then during testing they find something else, and then more of the same and so on.  Yes modern aircraft could easily be equipped to be unmanned.  Hell even older aircraft can do it, the military have been using F4 drones for some time.  Almost everything is ready to go, one could easily come to the conclusion that the only thing keeping them from doing it is regulation and legislation.

It’s no big trick to make a plane fly with no pilot; the .mil has done it with obsolete fighters for decades.  But that’s not what we are talking about.

Can you make it taxi out, take off, avoid weather, get to the destination, land, and taxi to the gate without human intervention, without additional infrastructure, and without any additional certification?   No, you cannot.   And doing that is what we are talking about.

1. Absolutely you can do this.

2. No that is not what we are talking about. That is what some of us are talking about. There are at least two, if not more, separate, different hypotheses in this topic. A) Total automation, no pilots at all, with executive level ground control of the aircraft. B) Single pilot op's supported by on-board automation.

(B) is already a fact of life in the Part 91 and 135 games. It appears inevitable for Part 121 cargo op's, especially with the advent of Garmin Autonomi Autoland. Lost contact with your pilot? Send an ACARS/FANS/whatever message to the aircraft to virtually push the Autoland button, pick up the phone and declare an emergency with the FAA. It's coming, folks.

That said, both human and underwriter (not undertaker ) pressures often force two pilot cockpits even when the aircraft and pilot are rated for single pilot op's. An easily accessible interweb example of that is Yahoo personality CitationMax. Lucky bastid got his dream career by becoming the pilot for the family business, which owns a CJ3+. Whenever he has "CitationDad" on board he is required to take on another pilot. So from that perspective I do agree that for Part 121 passenger op's it will be much longer, possibly never, for this.

At the end of the day you know who really draws the line. It's not the accountants, the FAA, NTSB or Garmin. It's the insurance underwriters. They are always the final arbiter of what does or does not happen.

Its fascinating that FedEx feels the need to be beta tester for this technology when their abysmal safety record in the 1990s and early 2000s nearly cost them their airline from insurance cold feet.

The liability is so immense that we will need legislation limiting damages nationwide before this could occur. Each death would be worth at least $7-10m each. Subsequent accidents would go for more due to allegations of wanton was since it was a "known" but "consciously disregarded" risk. That would put a plane load down in $1-2bn range - possibly more. Any pattern could drive the costs up at least 2-3x that. That's a massive pill for the market to swallow, and how many pilots could be bought for that.

The idea that pilots are super duper expensive, but networks are static and capable and fire-and-forget in the absence of expensive human touch labor is spurious.

If you ask the military if they are more afraid of a pilot shortage or a shortage of network folks, and which ones are harder to retain, they'll say the computer network people.

To me, for the military, the answer is amortized view of pilot training costs. For the civilian world, they get their pilot touch labor dirt cheap as is, as no major airline needs ab initio to fill classes.

No doubt. Just more reasons this is far fetched for 121.

This almost sounds like it's grant money being spent on a money loosing idea. Unless it's for entirely new aircraft development.

Isn't required flight crew a limitation of the type certificate?

Certification and regulatory as well.

The current "next gen" has taken 20 years to implement. Going unmanned will take something even more complicated to run, and even longer to implement.

And frankly even less secure than the current 121 environment, which most people would charitably describe at 4 parts theater to one part effective security.

Of course, the government is GREAT at hand-waving IT security.