SoylentNews is people
SoylentNews
A very small survey of people of different ages suggests that there are age and gender differences in the acceptance of riding in automated cars. In summary, 2,600 people in the US replied and of them 38% of the men and just 16% of women would be happy to ride in an automated vehicle. About a quarter of respondents said they would feel safe in a driverless car while around two thirds said they would not travel unless there was a driver. No mention was made about their opinions of sharing the road with these massive projectiles when driving themselves in traditional cars.
Source : Driverless cars: Men and women have very different opinions on letting go of the wheel
This discussion has been archived.
No new comments can be posted.
Men and Women Have Very Different Opinions on Automated Cars
|
Log In/Create an Account
| Top
| 136 comments
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
TOO BAD YOU CAN'T BUY a voodoo globe so that you could make the earth spin
real fast and freak everybody out.
-- Jack Handey, The New Mexican, 1988.
(Score: 3, Insightful) by Grishnakh on Friday February 02 2018, @01:05AM (7 children)
The code in aircraft and spacecraft is generally extremely reliable. And even the code in the engine controller or ABS system in your car is reliable; when was the last time you experienced a software error with those?
(Score: 2) by c0lo on Friday February 02 2018, @01:34AM (5 children)
Compare the R&D timeframe for a new airplane model with the timeframe for a new car model.
Compare the number of situations an airplane autopilot (or landing assist) has to handle with the number of situations a car "autopilot" needs to handle - eg tell me when the airplane autopilot needs to handle "kids playing ball on the footpath".
Compare the sensing equipment you can afford (in available space and weight) to mount on an airplane - phased radar array, radio equip - with the same for a car (some LIDARS, maybe some WiFi).
Compare the cost of infrastructure on an airport (radiobeacons, airstrip illumination/signalling) with the infrastructure of the same nature (semaphores, lane dividing lines, road signs) available on roads. Don't forget the maintenance state and related cost as well.
Do you really think I should continue enumerating the specific differences or are the above enough?
3 years back with the throttle control - overrevving or underrevving wildly - it went astray due to a defective sensor, needed to replace the sensor and the damned controller.
About $1800 and 4 weeks I couldn't use my car (fortunately those $1800 included the price for a courtesy car).
You may argue "that wasn't a bug". Maybe so, but it set into evidence one thing: no fallback solution from a failing electronic solution.
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by TheRaven on Friday February 02 2018, @12:09PM (2 children)
That's not a great comparison, because aircraft are so low volume that a lot of the components are bespoke for that model (or shared between a very small handful). In contrast, cars are high volume and high design turnover, so most 'car companies' buy huge amounts of off-the-shelf components from companies like Bosch. The R&D timeframes on these components are often several generations of car, but the end integrators just build with whatever the latest available parts are that year.
sudo mod me up
(Score: 2) by c0lo on Friday February 02 2018, @12:23PM (1 child)
Implicit assumption: the problem had already been solved (by Bosch and the like).
Reality check level 4 car autonomy hasn't been reached, timeline to level 5 (complete autonomy) is everybody's guess [techrepublic.com]
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 2) by TheRaven on Monday February 05 2018, @11:46AM
sudo mod me up
(Score: 2) by Grishnakh on Friday February 02 2018, @03:51PM (1 child)
3 years back with the throttle control - overrevving or underrevving wildly - it went astray due to a defective sensor, needed to replace the sensor and the damned controller.
About $1800 and 4 weeks I couldn't use my car (fortunately those $1800 included the price for a courtesy car).
What kind of car was this? And why would you need to replace the controller? Because the shop wants to make more money?
Compare the number of situations an airplane autopilot (or landing assist) has to handle with the number of situations a car "autopilot" needs to handle - eg tell me when the airplane autopilot needs to handle "kids playing ball on the footpath".
It can already handle that with the existing sensors. Also, if kids run out in front of cars, the driver is generally not at fault as long as they weren't speeding.
Compare the sensing equipment you can afford (in available space and weight) to mount on an airplane - phased radar array, radio equip - with the same for a car (some LIDARS, maybe some WiFi).
We have phased radar arrays on cars. Every car with "adaptive cruise control" has such an array on the front. That stuff has gotten seriously small and cheap. The arrays on aircraft are a lot larger and power-hungry though because they need far more distance ability (tens of miles, instead of maybe a quarter-mile at the very most).
Anyway, I agree that the complexity of driving is pretty daunting from a reliability perspective, when you're talking about the reliability of not hitting things and reacting properly to situations. I wasn't addressing that. I was addressing software reliability in the engineering sense: does it crash, is it buggy, etc.? Autonomous vehicle software can be perfectly reliable in executing its algorithms, while still not reacting that well to external situations and leading to a vehicle crash. The complaint before was that software is generally buggy, and I was pointing out that we already have lots of software that really isn't. Of course, you have a valid complaint about throttle-by-wire control software not handling a hardware fault gracefully, but even here this isn't evidence of "buggy" software, just a poor algorithm for handling hardware faults. As I asked before, when was the last time your engine controller blue-screened? This stuff just doesn't happen with properly designed embedded software.
(Score: 2) by c0lo on Friday February 02 2018, @04:54PM
The two are related. The more complex the situation the software has to face, the higher complexity of software.
The more complex the software, the higher chances of bugs.
And no, even when I used BSOD as an example, I wasn't restricting myself to "software crashes" - I used the example to show that any complex software will have bugs, some with fatal consequences. Especially commercially developed software, where the sales critters prevail in front of engineering.
Look, satellites are lost due to bugs and I have good reasons to believe the software they run is less complex than one needed for a level 4 or 5 car autonomy.
And really, I can't give a dam' if the software is so robust it doesn't crash and continue to run through its various bugs oblivious to consequences.
Mate, ABS software is so simple it can be implemented mechanically/hydraulically - it was certainly implemented this way in the '50-ies [wikipedia.org]. There can be no meaningful comparison between the complexity of ABS "software" and the one required by driveless cars.
Given the difference between a software that any decent microcontroller can run and a software to recognize and react properly to road situation, if you think the last will be bug-free, ...
Implicit assumption: the cars will have "properly designed embedded software".
Reality check - the auto industry couldn't design an entertainment system separate from the car control/navigation, which led to serious vulns. And this less than 3 years ago [wired.com]
https://www.youtube.com/watch?v=aoFiw2jMy-0 https://soylentnews.org/~MichaelDavidCrawford
(Score: 0) by Anonymous Coward on Friday February 02 2018, @03:40AM
> ABS system in your car is reliable; when was the last time you experienced a software error with those?
Last week, and every time I come to a snow-covered icy patch in a nearby mall, where the road is slightly downhill and then turns. These conditions are common here, all through the winter. The damn brakes make the usual noises (the ABS pulsing pump is working) and the car does not slow down to any useful extent. Luckily this car still has a manual hand brake that will lock the rear wheels and build up a wedge of snow that provides *some* deceleration. Obviously I now approach this spot slowly, and still wind up sliding through it--the sight lines are good enough that I can see if any other cars are going to be there and slow down even more in advance if needed.
The software is defective and when I've discussed with some of the authors (I work in a related auto engineering field) they throw up their hands and say that they haven't yet figured out how to deal with "deform-able surfaces".
How is your autonomous car going to deal with this? My guess is that it puts on the brakes and the ABS has it sliding right off the road.
Another car I use is old enough that it does not have ABS. This situation is easy to deal with -- just lock the wheels (very light brake pedal pressure since the surface is very slick). This builds up wedges of snow in front of *all four wheels* and the car slows down fairly well. Of course with wheels locked, it starts to slide and/or spin. After a second or two, I judge that it's time to release the brakes for a moment, long enough to do a little steering and get back on course. Then lock the wheels again. Repeat. No drama after a little practice.