What the Boeing Crashes can teach us about Stage 4 autonomous cars | FerrariChat

What the Boeing Crashes can teach us about Stage 4 autonomous cars

Discussion in 'Aviation Chat' started by bitzman, Mar 25, 2019.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, Skimlinks, and others.

  1. bitzman

    bitzman F1 Rookie
    BANNED

    Feb 15, 2008
    3,287
    Ontario, CA
    Full Name:
    wallace wyss
    I am trying to write an editorial about relating the Boeing crashes to the upcoming Stage 4 autonomous cars. I'll discuss it on a University radio program in Riverside. Even without the airline crash data fully analyzed, I think Boeing's lessons apply to the coming robot cars, any comments?
    ------------------------------------------------------------------------------------------------------------------


    Boeing Crashes:How We Can Apply What We Learned to Stage 4 Autonomous Cars

    When the Boeing 737 Max came out in 2017, it looked good. Better fuel efficiency. Airlines ordered them like crazy.

    Then came the crashes. The first in Indonesia, roughly 8 months ago. It was only after the second one, in Ethiopia, that it became clear that there was a pattern. The NYT published an article that showed the wild porpoise-like motions that both aircraft recorded, almost too much for a pilot to get ahead of even if he went to full manual.

    Now it’s coming out—that the foreign airlines buying the plane were told that it was just an updated version of the plane you already know, so it wasn't really necessary to buy simulator training on this version. So simulator training was left on the optional list and some airlines didn't order it.

    But even as crash results from the first crash are being analyzed, it is becoming known that the plane had a severe problem, the software would sense that it was going too slow so it would pitch the nose downward. When the pilot tried to bring it back up, it would fight the pilot and insist on diving again. Finally, after so many roller-coaster maneuvers, the pilot couldn’t keep up. And the plane went down. This is borne out by the NYT article that shows the altitude changes each plane went through, a frantic pitching up and down like a roller coaster.

    So it is coming down, even without full analysis of the second plane's black box, that it was a problem of pilot error--the pilots weren't equipped to handle and out-of-the-box problem.

    Max planes that were grounded may remain parked, and some of the 4000 orders Boeing had are being cancelled. It is a black eye for American industry.

    WILL THIS SAME SCENARIO HAPPEN WITH AUTONOMOUS CARS?

    We face a similar problem, involving potentially millions of drivers, with autonomous cars. At the present, we are only at Stage 2, the stage where there are optional lane departure warning devices, and in some cars, even automatic braking so you can drive in slow traffic and the car decides when to stop and start again to gain a few precious feet.
    But what the autonomous car lobby wants is Stage 5 by 2030. That is where there would be no steering wheel, no brake pedal. You get in the car, tell the robot brain where you want to go and it goes there.

    You can watch a video, play computer games, catch up on your sleep.
    But before Stage 5, there is Stage 4. The is where there is still a steering wheel, still a brake pedal. And still someone in the driver's seat who is supposed to be the "emergency safety driver. " That is to say, if the computers and radar or whatever doesn't react right to a situation, an audio signal cues the person in the driver's seat and they take over, going to manual. (which in the plane's accidents, might have saved the planes).
    But here is one problem with that. Some of the drivers may be extremely young, say 15 on a learner's permit, and some may be extremely old, over 80. And a lot more of them may be distracted by the video, the computer games, music selection, whatever they chose to entertain themselves while the robot handled the driving.
    You can visualize the scene; Mildred, age 78, is relaxing, knitting doilies as the Stage 4 car cruises at 70 mph. But then a sandstorm obliterates the white line marking the edge of the road. The car, once so confident, doesn't know where the road is anymore and drifts to the right, right off the pavement into the soft sand.
    Can Mildred, age 78, quickly drop whatever she is doing and wake up and grab the wheel and work the brakes? Can she, in one 250th of a second, adapt to a situation posed by bad weather that has wiped out the robot's usual navigational references? I think not.

    TRAINING ON CAR SIMULATORS FOR STAGE 4 DRIVERS
    We can learn from the Boeing disasters. What is needed before these cars go to the showrooms is training, training en masse, on simulators. We need everyone who is buying or renting a Stage 4 car, to sit at a simulator for a certain number of hours which will take them through what they will have to do to "take over control" if the robot loses his way.

    Oh, sure, I can hear the advocates of the autonomous car saying "But the robot won't lose his way. Robots don't have human frailties."
    Tell that to the two Tesla drivers whose Model S plowed into a white tractor trailer that cut across their path at an approximate 45-deg. angle. They were on what Tesla calls "Autopilot." The Tesla's radar didn't see the trucks.
    . We need thousands of simulators, and to determine how many hours at the wheel of the simulator each driver will need to be able to take over in an emergency. Some may, it turns out, feel after they try the simulator that they are not up to the task of taking over. So be it, they will be Uber or Lyft customers.

    Boeing has already come up with a software patch for the Max and is making some of the formerly-optional safety packages standard. Their mistake was thinking the pilots flying the plane, pilots from countries where few pilots own their own planes and are used to flying manually, all had enough manual flying experience to be able to take over in an emergency where the robot brain failed. They were wrong. Let's not make that same mistake with autonomous cars when we reach Stage 4. We are talking millions of cars that will be sold mostly on the Stage 4 attributes, a sort of "leave-the-driving-to-us" pitch.

    Stage 5, incredibly, is already being pushed at us, in some limited applications, like a van that picks up students and takes them to another part of the campus. But not in showrooms where they can be bought by anybody. Let's worry about perfecting Stage 4 first, or else we will see the Boeing disaster replicated one car at a time every time there's a circumstance that "blinds" the not-perfected-yet robots.....
     
  2. kylec

    kylec F1 Rookie
    Silver Subscribed

    Jun 9, 2005
    3,574
    Orlando
    Don’t. Please stop.
     

Share This Page