No Thank You

There seems to be a consensus that if driverless cars are ever to become universal, then the controlling system will have to be one single one — you can’t have competing, perhaps even incompatible systems fighting over the traffic. In other words, we’d need something like Europe’s Eurocontrol:

Eurocontrol’s Enhanced Tactical Flow Management System compares demand for flights in a particular area with the available capacity.
The system pulls together data such as flight plans, taxiing time, and flight position from numerous sources in multiple countries and collates them.
It can then track planes in real time to manage the number of planes in the air to make sure it doesn’t get too crowded.
Precise monitoring prevents the carefully balanced system from being thrown out by planes with delayed departures or arrivals.
Planes can then be herded into departure and landing slots at airports to keep the thousands of flights in Europe flowing smoothly.
ETFMS also helps plan flight schedules up to a week in advance to help airlines and air traffic controllers plan each day down to the minute.

Uh huh. Then something like this happens:

Up to half of all flights in Europe face delays today after a Europe-wide air traffic control system failed.
Eurocontrol, which runs the system, said that a technical problem means that as many as half a million passengers could be affected, disrupting travellers who went away for the Easter weekend.
‘Today 29,500 flights were expected in the European network. Approximately half of those could have some delay as a result of the system outage,’ Eurocontrol said. The agency said the system would be back up and running tomorrow.

The cause had been identified, it said, without saying what it was. The agency said ‘contingency procedures’ were in place to stop the system becoming overloaded but that these would be lifted later this evening.
Eurocontrol added that flight plans from before 11.26am BST were ‘lost’ and asked airlines to refile them.
The agency said it was a ‘technical fault’ and that the system had not been hacked, saying they were now ‘in recovery mode’.

“Lost”, huh? That’s comforting.

Here’s my takeaway. I am never going to submit myself to a driverless car. And I am certainly never going to board a pilot-less aircraft (which, incredibly, has been suggested by various airlines and aircraft manufacturers).

Systems fail occasionally — all systems fail eventually — and I’m not going to be a prisoner of this kind of happenstance, ever, if I can possibly avoid it.

19 comments

  1. Well Kim, if they can get an aerodynamic abortion like the V22 Osprey to fly… pilotless aircraft are practical. Without computers the V22 couldn’t fly at all. It will do that in hostile conditions too. A lot of men had to die to bring that about I suppose.
    I wonder if we aren’t seeing the forest for all the trees. We accept the fact that on any given night, the drunks will be on the road and behind the wheel. We now accept the fact that at least 2 out of every 10 drivers is an idiot or unqualified to operate a motor vehicle. Every year we scrape our kids off the pavement and guard rails after a drag race goes wrong.
    I kinda think the lawyers will handle this one for you and whether or not that is a good thing… is beyond me.

    1. If we learned anything from the passage of ObamaCare, it’s that we can’t trust the lawyers.

      1. Nor, as we should learn from the Obamacare debacle, can we any more trust actuaries. Because leftists desire it, it will come to pass, unless and until it becomes overwhelmingly demonstrated that it. Can. Not. Work. Set aside the potential body count. How many insurance companies will be bankrupted by ambulance chasers before the fools learn?

        A.I. doesn’t stand for artificial intelligence. It stands for aspirational idiocy.

        M

        1. I ask my passengers in illustration, “Your GPS app is an AI. When was the last time you thought it gave you the best possible routing over ground with which you are intimately familiar?

          M

    2. Speaking of lawyers, there’s another school of thought that says lawyers will kill the whole concept of self-driving cars. After all, who’s to blame in an accident? In your examples, it would be the driver – whether drunk, unqualified, or just young and stupid.

      In the automated future, it would be … ?

      Some programmer hundreds if not thousands of miles away, working on his/her/it’s 23rd hour of Red Bull fueled overtime trying to keep servers online and running?

      The manufacturer of the car? For not installing even more safety features?

      The spotty 3g reception caused by some electrical interference from good old mother nature?

      If there’s no one to sue, what are the lawyers to do?

      Finally, think of all the poor little Texas towns that are 90% financed by speed traps. What the heck will they do when every car on the road is electronically controlled to drive the exact speed limit all the time!

      1. As to the last, the moment that system is in place someone, somewhere, will start working on a way to hack into it to take over control of cars without leaving a fingerprint.
        And those corrupt police agencies will no doubt use that technology to make cars speed, then fine the drivers for it.
        Should be possible to leave a trace making it look like the driver hacked his car to allow him to drive himself, which no doubt will also be illegal at the time.

  2. I will never say never.

    As a fiercely independent person who observes the historical issues with central control, epitomized by government in general, I understand your position.

    However, as an aging person who has never liked driving and wants a rural lifestyle, I want to maintain my independence as long as I am physically mobile. Even when my reaction time or eyesight become too compromised to continue driving.

    In that regard, I only see 3 options, listed here in order of “most dependent” to “Most independent”: Public transportation (fail, doesn’t serve rural areas), Taxi/Lyft/Uber, and self-driving cars.

    1. Think of how many times people, trusting their GPS nav systems, have ended up in swamps or careening over cliffs, and you’ll see how flawed your third option is.
      Just last week, GPS wanted me to turn right onto a road which no longer exists — it had been a temporary construction road — and which had been closed off for TWO YEARS.
      My “driverless” car would have climbed the sidewalk and set off over a field.

      1. especially considering that updates to the satnav systems in cars tend to come at least 2-3 years after the situation changes. It just takes that long for the data to make it all the way through the process of being incorporated into the software updates and being made ready for distribution (most of which is spent waiting on government agencies to supply the data and then testing the product to see if it’s correct).
        And then you have to wait for your actual system to get updated in your actual vehicle…

  3. I too recognize the risk that Kim is pointing out; and I have all manner of reservations about self-driving cars. If I lived in a metro area, I’d be all-in on any campaign to keep them off the streets. (well, at least the local streets, see below)

    However, as a resident of rural America, I agree with lpdbw (even without considering old-age related mobility).

    It takes four to five hours of drive time from where I live to my state’s capital. If there were a self-driving car that could do that trip — letting me read, work, watch a movie, or just look at the sights on the way — I’d buy one this afternoon.

    PS – the geniuses working on self-driving cars are doing it wrong. They are trying to solve automated driving in urban, local-street conditions. If they focused instead on making a car that did the driving JUST on the freeway, but handed things off to the driver when you got close to the off-ramp, they’d:
    a) have an easier problem to solve, in terms of artificial intelligence, etc., so they could get a working vehicle to market faster.
    b) be able to begin showing that this stuff works in the real world, and establishing whether or not society could function with automated vehicles in the mix. (Because, as Kim notes, if it takes true centralized command and control of every vehicle on the road before this works, then it ain’t ever gonna happen).
    c) have a bigger market than they suppose, because it’s not just rural deplorables like myself who’d use such a vehicle. Even in urban areas, the vast majority of commuters spend some portion of their trip on a limited-access freeway. Nearly every driver in the country would have the opportunity to use the feature — daily.

    1. Very good point. It would certainly make my daily commute through the hell of Seattle area traffic easier. (Thank God for podcasts…)

      Not that I’d be sitting there reading on the freeway, I program machines and robots for a living and do *not* trust systems that far, but, being able to ever so slightly relax would be nice.

      An intermediate step would be a new truck I’m eyeing that has adaptive cruise control.

    2. Keep in mind that a LOT of the effort here is going into automating *trucking*. It’s a YUGE market where corporations are EAGER to fire…I mean replace…I mean are EAGER to get self driving trucks on the road where they will increase safety and decrease costs for the corp…I mean everyone.

      Oh, and put 10s of thousands of deplorable type truck drivers out of work.

  4. I am well into my 70’s and I enjoy driving, here in the Texas Hill Country (yep it’s a thing) we have some great up and down curvy roads that make things interesting. People come from all over to drive their motorcycles in this area and then end up in Luckenbach 35 miles from us so they can drink beer and look cool, or something.

    I also don’t mind long road trips, last year to Jacksonville FL and Savannah GA and next year back to Colorado where the roads really do go up and down, I know how long I can drive and I have enough sense to stop and rest when need be. I see a problem not only with self driving cars but with every damn other labor saving gadget and wonder if the time will come when people with some means won’t have to do anything but go to a gym to workout.

  5. Hmmm!
    self-driving cars in a rural environment.
    didn’t those used to be called horses at some point
    no good support system for them now

  6. > There seems to be a consensus that if driverless cars are ever to become universal,
    > then the controlling system will have to be one single one

    I don’t know who is touting this “consensus”, but I doubt it’s even as real as the “consensus” regarding global warming.

    It would be easier and safer to build a self-driving car without central control.

    There will be (and self driving cars WILL happen) standards and protocols that these cars have to work on, but anyone who says that these things have to be “centralized” is…they have another agenda.

    > My “driverless” car would have climbed the sidewalk and set off over a field.

    An driverless car safe enough to test on the streets would have recognized that there was no road there.

    I don’t really *want* a driverless car right now, and if they never happen as I age I’ll adapt my lifestyle to deal with less ability to drive. But if I had to commute “downtown” in the sort of traffic that is common these days I’d buy one as a third car IN A FLIPPING SECOND. A small two seater electric with just enough trunk space for a medium sized suitcase would be fine for that.

    (And yes, I know that last sentence made Kim throw up in his mouth a little :), but it’s not a *lifestyle* car, it’s a tool to get me to my cubical and home. I’ll still have a 4WD SUV for getting up into the mountains whenever I want).

  7. My smartphone and lappy is pretty mature tech, and I still have to restart the things every couple of weeks because they get cranky.
    I’ve got various bits of satellite tech I use, and they still go out because of thunderstorms.
    People got to be careful not to get so excited about the potential benefits of a proposed thing that they overlook the possible bad sides. Because someone will be looking hard to find the bad sides, and use it against you.

  8. I think it’s going to be a long time before fully automated cars become standard. There will for a long time be a need for a manual override. Computers just aren’t that clever right now.

    What I do see is automated vehicles kicking in is for selected routes like motorways / interstates where the necessary infrastructure can be more easily installed. That is, you drive to the motorway and hand over to the computer. I’ve heard about this being planned for fully automated lorries. They plan for depots next to the motorways and the lorries are to drive between the depots.

  9. I’ve been in the business of testing unmanned aircraft since 1995. Building a self-driving car is feasible, but challenging. Stay clear of central control, it’s too damned vulnerable.

    What troubles me is that you have software companies playing this game – and they have NO clue how much testing is required to ensure safety of operation. Hell, Microshoddy ran into problems with Windows 10 killing Word…their own damned software! On a moving vehicle, “Blue screen of death,” isn’t a figure of speech. Somebody just died.

Comments are closed.