Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Pretty Cool video from Bosch - Automated Driving in a MS

This site may earn commission on affiliate links.
This is a very cool video and excites me about the future of automatic driving in my Model S. However, I think Phillip is a way too distracted when the car is in automatic mode. Maybe I'm being a bit too paranoid about the possible failings of technology but watching videos and reading email seems a little too trusting to me. If/when my Model S is fully autopilot capable on the highway, I intend to still pay full attention to what's going on.
 
This is a very cool video and excites me about the future of automatic driving in my Model S. However, I think Phillip is a way too distracted when the car is in automatic mode. Maybe I'm being a bit too paranoid about the possible failings of technology but watching videos and reading email seems a little too trusting to me. If/when my Model S is fully autopilot capable on the highway, I intend to still pay full attention to what's going on.

If you can't detach long enough to read an email, what's the point of automatic driving?

Eventually, you should be able to take eyes off the road long enough to read, sleep, rotate the driver chair to face the rear, etc.. That's what I would expect from "automatic driving"
 
This is a very cool video and excites me about the future of automatic driving in my Model S. However, I think Phillip is a way too distracted when the car is in automatic mode. Maybe I'm being a bit too paranoid about the possible failings of technology but watching videos and reading email seems a little too trusting to me. If/when my Model S is fully autopilot capable on the highway, I intend to still pay full attention to what's going on.

+1000%

No way even the most advanced "automated driving" computer can account for sudden and erratic movements from other drivers, or small to medium hazards in the roadway. I can think of a dozen examples of how this (as demonstrated in the video) would quickly fail causing a potentially fatal accident if the driver is as distracted as "Phil".

Automation == Good. Removing the driver entirely=Bad. Very, Very, Bad.

* Unless every car on the road is similarly automated and *all* humans are removed from the equation. That still doesn't protect against road hazards.
 
Like this overall. Would be unsurprised if the "top view" (what I refer to as God view) at 43 and 58 secs in, is possible today as alluded to in 7.0 update.

Seems that Bosch has been actively talking about " a range of driver-assistance technologies, including adaptive cruise control, ESP, lane-detection warning, radar and ultrasound sensors and multi-purpose video cameras." for some time, eg Bosch sees gradual progression to autonomous driving - automotiveIT International
 
+1000%

No way even the most advanced "automated driving" computer can account for sudden and erratic movements from other drivers, or small to medium hazards in the roadway. I can think of a dozen examples of how this (as demonstrated in the video) would quickly fail causing a potentially fatal accident if the driver is as distracted as "Phil".

Automation == Good. Removing the driver entirely=Bad. Very, Very, Bad.

* Unless every car on the road is similarly automated and *all* humans are removed from the equation. That still doesn't protect against road hazards.

Automation doesn't need to be perfect, it just needs to be better than human drivers.

Google cars have already been proven to be better than human drivers, even in city streets.

What cars lack in natural intuition, they make up for with blazing-fast and incredible response time, an always-watching approach 24/7, the ability to see 360 degrees in real time at all times, the ability to see in the dark and see through weather conditions, and they will never break the law internationally or be in a hurry.
 
Automation doesn't need to be perfect, it just needs to be better than human drivers.

Google cars have already been proven to be better than human drivers, even in city streets.

I don't think automated drivers will *ever* be better than human drivers in our lifetimes, unless, like I said, humans are removed from the equation and all cars are automated on the highway.

I don't believe that the Google cars have been "proven" to be better than human drivers. I'm not sure how that's even possible unless you define "better" with some extremely narrow criteria.

Also, from what I've seen, they work on city streets, but very, very, slowly.

Automated drivers can probably take care of 98% of the routine driving. But it's that pesky 2% they'll never be able to handle, and that's where fatal accidents will happen.
 
If you can't detach long enough to read an email, what's the point of automatic driving?

Eventually, you should be able to take eyes off the road long enough to read, sleep, rotate the driver chair to face the rear, etc.. That's what I would expect from "automatic driving"

Just wondering what your best guess is how many years before "eventually" happens? 5 years, 10 years, 20 years?? It should also be noted that Elon is very careful to use the phase autopilot and not automatic driving. There is a significant difference especially if you expect to be able to "sleep". I'll pass on that feature:smile:
 
I think that in less than 10 years automated driving will be demonstrably safer than human driving in most circumstances. Severe weather will remain a challenge for automated systems, but of course it is also a challenge for humans.

That video was shot on the 280 freeway a few miles northwest of my house. Of course the scenes were out of order from a geographic point of view, but no one would notice unless they were very familiar with that freeway.
 
?..
I don't believe that the Google cars have been "proven" to be better than human drivers. I'm not sure how that's even possible unless you define "better" with some extremely narrow criteria...

Build 2000 self-driving cars. Give them to randomly chosen drivers half with the automation turned on and the drivers instructed never to touch the controls except NAV, the other half with the automation off. And the drivers instructed to simply drive as usual. After a year see which group has fewer and less serious accidents.
 
Good experiment suggestion. Also instead of wondering if self-driving cars will be better than you, ask yourself if self-driving cars will be better than the person in the car next to you. Ego and cognitive bias hinder an answer to the first question so the answer to the second question is more likely to be accurate.
 
Good experiment suggestion. Also instead of wondering if self-driving cars will be better than you, ask yourself if self-driving cars will be better than the person in the car next to you. Ego and cognitive bias hinder an answer to the first question so the answer to the second question is more likely to be accurate.

+1

I think a computer could beat me today and most certainly the idiot next to me. I don't know where all this "never" comes from.

Most accidents are a result of distracted or drunk drivers. A computer is never either of those things.

As always these threads get into pro-autonomous driving and anti-. This being a car forum is over represented with hard core drivers. But also over represented with techies. So both extremes.

I love to drive but I hate the wasted time/attention more than I love to drive.

I believe the federal government is asleep at the wheel here (as most things) and needs to start the process of initiating standards for autopilot and beyond. If manufacturers have rules and standards, it protects them from the legal issues. How we don't have a car to car communication protocol yet is ridiculous. Driving on 95 in NC yesterday showed the folly of human drivers. Going north (other side of road) there was plenty of room to travel 70 mph safely but there were repeated full on stops because of driver anxiety. The wasted time and energy was disgusting to watch.

For all those "drivers" - just look at the wasted time, energy, loss of life, and up to $800 billion a year in economic damages. Is it really worth it?
 
+1

I think a computer could beat me today and most certainly the idiot next to me. I don't know where all this "never" comes from.

Most accidents are a result of distracted or drunk drivers. A computer is never either of those things.

As always these threads get into pro-autonomous driving and anti-. This being a car forum is over represented with hard core drivers. But also over represented with techies. So both extremes.

I love to drive but I hate the wasted time/attention more than I love to drive.

I believe the federal government is asleep at the wheel here (as most things) and needs to start the process of initiating standards for autopilot and beyond. If manufacturers have rules and standards, it protects them from the legal issues. How we don't have a car to car communication protocol yet is ridiculous. Driving on 95 in NC yesterday showed the folly of human drivers. Going north (other side of road) there was plenty of room to travel 70 mph safely but there were repeated full on stops because of driver anxiety. The wasted time and energy was disgusting to watch.

For all those "drivers" - just look at the wasted time, energy, loss of life, and up to $800 billion a year in economic damages. Is it really worth it?

Automated driving will have as big an impact on the 21st century as cars did on the 20th. We have about 70 years of urban planning about to be made obsolete (over 20-30 years) as autonomous cars will reduce the number of cars in use and the space needed for roads and parking. I'm 47 now and up until about three years ago I wondered if autonomy would happen in my lifetime, now it is inevitable and coming much faster than our society is equipped to react.

The politicians and the law will be creeping along well behind the technology and popular demand (as usual) but they will eventually respond, just as they did to cars in the first half of the 20th century.

No, cars will still not fly! The idea of maintaining the independence of mobility that an autonomous car can provide the old and the disabled is much better.
 
Just wondering what your best guess is how many years before "eventually" happens? 5 years, 10 years, 20 years?? It should also be noted that Elon is very careful to use the phase autopilot and not automatic driving. There is a significant difference especially if you expect to be able to "sleep". I'll pass on that feature:smile:


The sooner, the better.

Google self driving cars being tested today don't even have a steering wheel. Without driving controls it doesn't matter what the human in the car is doing.
 
I don't think automated drivers will *ever* be better than human drivers in our lifetimes, unless, like I said, humans are removed from the equation and all cars are automated on the highway.

I don't believe that the Google cars have been "proven" to be better than human drivers. I'm not sure how that's even possible unless you define "better" with some extremely narrow criteria.

Also, from what I've seen, they work on city streets, but very, very, slowly.

Automated drivers can probably take care of 98% of the routine driving. But it's that pesky 2% they'll never be able to handle, and that's where fatal accidents will happen.

It's already proven, though, unless you don't consider accident rate and fatality rate a worthy statistic indicative of safety.

Compare the accident rate of Google vehicles on the highway to the accident rate of human drivers on the highway. Both statistics are readily available. We actually do not yet know how much safer automated highway driving is at this point, because a Google vehicle has never yet been involved in an accident while in autopilot mode despite millions of miles being driven. The same is actually true of autopilot in city driving as well, but we only have hundreds of thousands of miles there rather than sever

You could make the argument that they're not practical for all situations yet, but to say that autopilot is less safe than the average human driver is demonstrably false in 2015.