Linked by Thom Holwerda on Thu 19th Jan 2017 21:47 UTC
In the News

The administration's analysis of Autosteer was more positive about its capabilities. After analyzing mileage and airbag deployment data for Model S and Model X cars equipped with Autopilot, the NHTSA concluded that "the Tesla vehicles' crash rate dropped by almost 40 percent after Autosteer installation."

Wait, you mean to tell me a computer who doesn't get sleepy or distracted and doesn't need to pee is better at keeping an eye on the road than a human?

Say it isn't so.

Order by: Score:
Perspective
by Alfman on Fri 20th Jan 2017 03:03 UTC
Alfman
Member since:
2011-01-28

We need to factor in how typical humans will react to this technology as it becomes more pervasive; from the article:

Although the NHTSA probe was looking for defects in the system, the more serious concern was that Tesla was giving its drivers too much confidence in a system that still requires constant driver attention. In reviewing how Tesla warns it drivers about the need for attention to the road, the NHTSA took into account a software update the company pushed in September 2016 that adjusted how quickly the car would warn the driver if their hands were not on the wheel. The September update also began timing out Autopilot if the driver failed to respond to the car's warnings promptly.

Reply Score: 3

Comment by Meor
by Meor on Fri 20th Jan 2017 06:01 UTC
Meor
Member since:
2006-09-29

Autonomous vehicles will revolutionize transportation and safety.

The only thing that could stand in its way is government regulations.

Reply Score: 2

RE: Comment by Meor
by kwan_e on Fri 20th Jan 2017 06:41 UTC in reply to "Comment by Meor"
kwan_e Member since:
2007-02-18

Autonomous vehicles will revolutionize transportation and safety.

The only thing that could stand in its way is government regulations.


Or more likely regulations will help make ubiquitous autonomous vehicles a reality. Everything from mandating that roads, vehicles and the surrounding environment be made easier for autonomous vehicles; to making autonomous vehicles mandatory.

I see more opposition to autonomous vehicles from people who would rather be responsible for killing people than let go of full control of their cars at all times. They love dreaming up rare scenarios where autonomous vehicles MIGHT fail.

Reply Score: 4

RE[2]: Comment by Meor
by boofar on Fri 20th Jan 2017 07:13 UTC in reply to "RE: Comment by Meor"
boofar Member since:
2008-04-23

They love dreaming up rare scenarios where autonomous vehicles MIGHT fail.


You mean, like, roads and signs covered in snow? Yeah, that'll never happen.

Reply Score: 3

RE[3]: Comment by Meor
by kwan_e on Fri 20th Jan 2017 07:38 UTC in reply to "RE[2]: Comment by Meor"
kwan_e Member since:
2007-02-18

"They love dreaming up rare scenarios where autonomous vehicles MIGHT fail.


You mean, like, roads and signs covered in snow? Yeah, that'll never happen.
"

Compared to humans? Autonomous vehicles would probably outperform humans in those areas too if they were developed to handle those situations.

Reply Score: 3

RE[3]: Comment by Meor
by benoitb on Fri 20th Jan 2017 08:36 UTC in reply to "RE[2]: Comment by Meor"
benoitb Member since:
2010-06-29

Provided there is an accurate road database (there should be if you approve use of these cars ?) the autonomous cars should know about signs covered in snow better than a human driver does.

Reply Score: 3

RE[4]: Comment by Meor
by modmans2ndcoming on Mon 23rd Jan 2017 00:23 UTC in reply to "RE[3]: Comment by Meor"
modmans2ndcoming Member since:
2005-11-09

heck, if autonomous cars are pervasive then the AI can simply record the data about the roads from each car and submit variances to the database online, this will mean even if something is covered in snow it will know what is supposed to be there (stop sign, pot hole, construction, etc)

Reply Score: 2

RE[3]: Comment by Meor
by unclefester on Sat 21st Jan 2017 08:13 UTC in reply to "RE[2]: Comment by Meor"
unclefester Member since:
2007-01-13


You mean, like, roads and signs covered in snow? Yeah, that'll never happen.



The solutions to these problems are actual very simple - embedded magnetic wires to mark lanes and road edges and passive radio transponders installed in road signs.

Reply Score: 3

RE[4]: Comment by Meor
by modmans2ndcoming on Mon 23rd Jan 2017 00:24 UTC in reply to "RE[3]: Comment by Meor"
modmans2ndcoming Member since:
2005-11-09

No need for that. It would make autonomous cars less flexible.

Reply Score: 2

RE[2]: Comment by Meor
by jgfenix on Fri 20th Jan 2017 08:31 UTC in reply to "RE: Comment by Meor"
jgfenix Member since:
2006-05-25

Autonomous cars will perform well in roads designed (or adapted) for them. There are roads only apropiate for goats where I wouldn´t trust an autonomous car in a thounsand of years.

Reply Score: 3

RE[3]: Comment by Meor
by kwan_e on Fri 20th Jan 2017 08:58 UTC in reply to "RE[2]: Comment by Meor"
kwan_e Member since:
2007-02-18

There are roads only apropiate for goats where I wouldn´t trust an autonomous car in a thounsand of years.


That's right now. Projecting into the future normally shouldn't be "what happens if we freeze technology at today's state and only move time forward".

Reply Score: 3

RE[3]: Comment by Meor
by Alfman on Fri 20th Jan 2017 10:02 UTC in reply to "RE[2]: Comment by Meor"
Alfman Member since:
2011-01-28

jgfenix,

Autonomous cars will perform well in roads designed (or adapted) for them. There are roads only apropiate for goats where I wouldn´t trust an autonomous car in a thounsand of years.



There are undoubtedly some conditions where the AI risk rises disproportionately. Ideally the technology could recognize this and explicitly inform the user that it is having trouble. This way the human driver doesn't place more confidence in the AI than the AI does in itself.

It's statistically implausible that the device will have 0% error rate, and so it's only a matter of time before one causes a crash, even if only "one in a million". But with relatively worse human accident rates it could still be worth the tradeoff.

Reply Score: 4

RE[4]: Comment by Meor
by dsmogor on Fri 20th Jan 2017 16:47 UTC in reply to "RE[3]: Comment by Meor"
dsmogor Member since:
2005-09-01

That's not gonna work in long term for the following reasons:
0. The AI enabled roads will never have 100% coverage due to too high cost/benefit ratio for certain areas (not necessarily within the US)
1. If critical mass AI enabled roads on given area is achieved large number of drivers will stop driving for prolonged periods.
2. Their skills will deteriorate to the point of uselessness or at least increased risk on non-AI roads.
3. Govt will have to threaten them with revocation of their licenses
4. Faced danger of loosing ability to drive a car outside of AI-enabled areas (which at that point will be large but not fully sufficient) lots of drivers will choose to abandon self driving cars

Reply Score: 2

RE[5]: Comment by Meor
by unclefester on Sat 21st Jan 2017 08:24 UTC in reply to "RE[4]: Comment by Meor"
unclefester Member since:
2007-01-13

That's not gonna work in long term for the following reasons:
0. The AI enabled roads will never have 100% coverage due to too high cost/benefit ratio for certain areas (not necessarily within the US)
1. If critical mass AI enabled roads on given area is achieved large number of drivers will stop driving for prolonged periods.
2. Their skills will deteriorate to the point of uselessness or at least increased risk on non-AI roads.
3. Govt will have to threaten them with revocation of their licenses
4. Faced danger of loosing ability to drive a car outside of AI-enabled areas (which at that point will be large but not fully sufficient) lots of drivers will choose to abandon self driving cars


Are you aware that airline pilots typically spend less than one minute at the controls during a flight? Pretty much everything except the takeoff and landing is totally automated. The technology already exists to completely eliminate the pilots.

Reply Score: 3

RE[5]: Comment by Meor
by Drumhellar on Sat 21st Jan 2017 19:21 UTC in reply to "RE[4]: Comment by Meor"
Drumhellar Member since:
2005-07-12

0. The AI enabled roads will never have 100% coverage due to too high cost/benefit ratio for certain areas (not necessarily within the US)


So? This is already the case with "Honda Civic" enabled roads, or "Ford F250" enabled roads. No matter the vehicle you buy, there will be some sort of limitations on certain roads that will make your vehicle a poor choice.

Reply Score: 2

Just noticed
by kwan_e on Fri 20th Jan 2017 08:56 UTC
kwan_e
Member since:
2007-02-18

The title says "Tessla".

Reply Score: 2

It will change nothing essential
by lighans on Fri 20th Jan 2017 11:22 UTC
lighans
Member since:
2006-01-14

As long as self driving cars are cars, nothing will change in traffic. The view from a city with normal cars and self driving cars are the same.

Real change would come if people would get out of there car, sell that thing and buy a bicycle. The change that this driver will kill another person is zero. And that will never be for cars (with or without assistance).

It is in general a strange idea to choose for a 2000 kg heavy vehicle generating poisonous gas to go to the grocery a few kilometres further. And it is still a strange idea if that vehicle is electric.

If we want real change, get out of your car.

There are numerous of studies about cities without access of cars, where more stuff is sold than similar city with car access. Also many people realize that their bicycle is faster in rural areas then cars.

In my view making self driving cars is a distraction from the real problem. The car itself.


(I have to admit that I am writing this in a very bicycle friendly country)

Reply Score: 1

kevstev Member since:
2011-07-14

I was with you until this:

"Also many people realize that their bicycle is faster in rural areas then cars."

On what basis are you making that claim? I could support the reverse claim- I can get around NYC faster on Citibike than I ever could if I wanted to drive my car places. Finding parking is a very time consuming process, and biking isn't a whole lot slower than even taking a cab.

However, once I get out of highly urban areas, where you can drive 30mph+ and the distances become larger and parking is plentiful, driving is far faster for all scenarios.

Reply Score: 4

lighans Member since:
2006-01-14

Exactly. So one could use a bike in rural areas and a car for the country.

Reality is that the car is only used. And a Tesla will not change that.

Reply Score: 1

Alfman Member since:
2011-01-28

lighans,

Real change would come if people would get out of there car, sell that thing and buy a bicycle. The change that this driver will kill another person is zero. And that will never be for cars (with or without assistance).

It is in general a strange idea to choose for a 2000 kg heavy vehicle generating poisonous gas to go to the grocery a few kilometres further. And it is still a strange idea if that vehicle is electric.

If we want real change, get out of your car.
...
(I have to admit that I am writing this in a very bicycle friendly country)


There are times I really like the idea of not needing a car. But as you hinted, unless you live in a city optimized for foot/bicycle traffic with local commerce and jobs, then the lack of a car introduces some major logistical challenges for employment, shopping, and even child care. The reality is that many of us are dependent on regular long distance travel for work, chores, food, school, child car, visit relatives, doctors, and other basic necessities in modern times.

I think there's been a cyclic feedback loop as cars eliminate the need for local mom & pop stores, which have been gutted and replaced by superstores covering a much larger area and are further away. This increases our need for vehicles for both shopping and employment.

Theoretically we might reverse the mass consolidation of businesses and reintroduce local stores to serve foot traffic again, which could create more jobs at the same time. But the truth is that the small local businesses are struggling so badly in the first place due to the economic pressures of capitalism to consolidate.

Make no mistake, I think it really sucks that companies expect people to commute so far for everything, but that's the modern reality many of us are facing. Eliminating cars would require solutions to the capitalistic forces that produce consolidation. How do we compel businesses to open up small local stores again given that de-consolidation would necessarily eat into corporate profits? While it may not be a bad thing to redistribute wealth to local economies, it is still a political quagmire.

Reply Score: 3

lighans Member since:
2006-01-14

That is correct.

This weekend I had a tour through Rotterdam. A "big" city in NL. There was a huge market which you can reach by car but much easier by bike. It was filled with people.

I think that an infrastructure which stimulates moving with a bike or walking, instead of sitting in a car would have big benefits. There are many examples. there is an American school which forces parents to bring their children by foot or bike. And it works. The school is growing. One of many examples.

Reply Score: 1

RE: It will change nothing essential
by JLF65 on Fri 20th Jan 2017 15:59 UTC in reply to "It will change nothing essential"
JLF65 Member since:
2005-07-06

People are killed by bikes every year. Not as many as cars, but it's not zero, either. Not to mention as long as there are cars and trucks and no special bike lanes, you take your own life in your hands to try to share the road on a bike. Far more people are killed ON bikes by cars and trucks.

Even given special paths to stay safe, there are many times a bike won't be any use: bad weather, need lots of carrying capacity, have a large family you can't leave on their own, have a long distance (more than a few miles/kilometers), have really steep terrain, etc. There's a reason that cars and trucks dominate travel the world round. Bikes are a shrinking minority other than a scant few places.

Reply Score: 2

lighans Member since:
2006-01-14

Far more people are killed ON bikes by cars and trucks.


So let's blame the victim?

Even given special paths to stay safe, there are many times a bike won't be any use: bad weather, need lots of carrying capacity, have a large family you can't leave on their own, have a long distance (more than a few miles/kilometers), have really steep terrain, etc. There's a reason that cars and trucks dominate travel the world round. Bikes are a shrinking minority other than a scant few places.


I think you are talking about an American situation?
My family of 4 has 10 bikes. A cargo bike for shopping or bringing the kids. Up to 4 kids and a lot of groceries can be carried. Some normal bikes for kids (and their BMXes). A fully enclosed recumbent bike (velomobile) for my travel to work. And some other.

Every bike has his specialism. We really need more space in the garage. ;)

Bad weather? Did you actually count how many times you have to drive in the rain? It is below 10%. There is no bad weather, but their is a lot of fear for bad weather. Bigger problem is smog.

Reply Score: 1

RE: It will change nothing essential
by mkone on Fri 20th Jan 2017 18:21 UTC in reply to "It will change nothing essential"
mkone Member since:
2006-03-14

As long as self driving cars are cars, nothing will change in traffic. The view from a city with normal cars and self driving cars are the same.

Real change would come if people would get out of there car, sell that thing and buy a bicycle. The change that this driver will kill another person is zero. And that will never be for cars (with or without assistance).

It is in general a strange idea to choose for a 2000 kg heavy vehicle generating poisonous gas to go to the grocery a few kilometres further. And it is still a strange idea if that vehicle is electric.

If we want real change, get out of your car.

There are numerous of studies about cities without access of cars, where more stuff is sold than similar city with car access. Also many people realize that their bicycle is faster in rural areas then cars.

In my view making self driving cars is a distraction from the real problem. The car itself.

(I have to admit that I am writing this in a very bicycle friendly country)


There is nothing "strange" about wanting comfort and convenience, particularly when there is no easy way to get to and from the shops if you are going to buy a sizeable quantity of stuff. A few kilometres away might be an hour or more of walking in either direction. In some locations, a car can do that in 5 minutes.

In cities and towns optimised for cars (like many American cities / suburbs), it is no surprise to see people take the car to the shops. It is also not silly or strange.

Reply Score: 2

kwan_e Member since:
2007-02-18

There is nothing "strange" about wanting comfort and convenience, particularly when there is no easy way to get to and from the shops if you are going to buy a sizeable quantity of stuff. A few kilometres away might be an hour or more of walking in either direction. In some locations, a car can do that in 5 minutes.

In cities and towns optimised for cars (like many American cities / suburbs), it is no surprise to see people take the car to the shops. It is also not silly or strange.


I think his point (I haven't read any studies he alluded to) is that the "comfort" and "convenience" afforded by cars actually works against the aim of getting as much people into a city.

Yes, a person can buy more stuff if they had a car, but a person cannot buy more stuff than, say, four people buying some stuff. Most people don't fully load up the car every time they buy something, so there is something to be said about making things more convenient for frequent small purchases.

And cars don't scale well because car parks don't scale well. Car parks may make a lot of money for their owners, but that takes money away from the retailers, who would arguably benefit the local economy more if people spent money on them rather than on parking.

Reply Score: 3

lighans Member since:
2006-01-14

There is nothing "strange" about wanting comfort and convenience, particularly when there is no easy way to get to and from the shops if you are going to buy a sizeable quantity of stuff. A few kilometres away might be an hour or more of walking in either direction. In some locations, a car can do that in 5 minutes.

In cities and towns optimised for cars (like many American cities / suburbs), it is no surprise to see people take the car to the shops. It is also not silly or strange.


Do you realize that people are going by car to the fitness centre to stay healthy?

I think that is the contradiction we are living in. We want to be healthy persons and make the wrong choices.

Cities optimised for cars force you to take a car. But biking that big lanes is very easy.

Comfort and convenience is great. But honestly. When I drive to my work when the some is coming up, the wind through my hair, birds, cows and waving to other bikers who are commuting. That is great. Al lot more convenient then having to look at the grumpy quickly annoyed cardrivers. ;)

Reply Score: 1

unclefester Member since:
2007-01-13

As long as self driving cars are cars, nothing will change in traffic. The view from a city with normal cars and self driving cars are the same.

Real change would come if people would get out of there car, sell that thing and buy a bicycle. The change that this driver will kill another person is zero. And that will never be for cars (with or without assistance).

It is in general a strange idea to choose for a 2000 kg heavy vehicle generating poisonous gas to go to the grocery a few kilometres further. And it is still a strange idea if that vehicle is electric.

If we want real change, get out of your car.

There are numerous of studies about cities without access of cars, where more stuff is sold than similar city with car access. Also many people realize that their bicycle is faster in rural areas then cars.

In my view making self driving cars is a distraction from the real problem. The car itself.


(I have to admit that I am writing this in a very bicycle friendly country)


I live in a very hot and humid subtropical city with torrential summer rain and very steep hills. You would find it very difficult to rely on a bicycle for transport.

Reply Score: 2

przemo_li Member since:
2010-06-01

There will always be a need for motorized delivery. Stuff is either too big or weather too severe.

Autonomous (and electrified) then is best option.

Autonomous because vehicle occupancy can be maximized, and number of vehicles on the roads minimized.

Electrified for moving air pollution outside city borders.

Reply Score: 3

Well...
by MeTitus on Fri 20th Jan 2017 13:08 UTC
MeTitus
Member since:
2011-08-19

I find it so crazy how people can easily be taken to accept everything which makes them even more lazy. No I do not want to my life to be driven by technology and yes I still like to drive. The way it is going, it looks a lot of people would just prefer to stay at home doing nothing all day and be told what to do even, by AI. These world is doomed.

Reply Score: 1

RE: Well...
by JLF65 on Fri 20th Jan 2017 16:07 UTC in reply to "Well..."
JLF65 Member since:
2005-07-06

A lot of the time, it's not lazy at all - it's busy or distracted. That businessman driving to work while trying to shave, finish a report, talk on the phone, and check his text messages. He NEEDS an automated car. There are many others like him on the road where all of us would be safer if they weren't also in "control" of the vehicle.

Well, ideally he would take mass transit, but that's not gonna happen in the US. Unless you live someplace like NYC, mass transit is a joke. We are more likely to see services like Uber or Lyft with automated vehicles that go around taking folks to work than to see meaningful mass transit.

Reply Score: 2

RE: Well...
by modmans2ndcoming on Mon 23rd Jan 2017 00:29 UTC in reply to "Well..."
modmans2ndcoming Member since:
2005-11-09

I want to be able to be productive on my commute to work. I want to be able to take a road trip with my family and engage with them on the drive rather than yell about how they need to shut up so I can concentrate.

Reply Score: 2

It isn't so
by Carewolf on Sat 21st Jan 2017 00:19 UTC
Carewolf
Member since:
2005-09-08

See topic

Reply Score: 2

A joke
by przemo_li on Sat 21st Jan 2017 12:43 UTC
przemo_li
Member since:
2010-06-01

There is this joke about IT.

"""
Director comes to senior developer and ask for rough time&space estimation for an idea he have:

* We need an mobile application.
Dev: OK. We have 2 people for the job right now.

* App will track user movement.
Dev: No, problem. 2 hours.

* App will know if user is in natural park
Dev: No, problem. 2 days.

* App will allow taking photos.
Dev: No, problem. 4 days.

* App will detect bird on that photo.
Dev: Ok. I will need 6 people and budget for R&D for 2 years.
"""

Reply Score: 3

Comment by schrepfler
by schrepfler on Sun 22nd Jan 2017 17:35 UTC
schrepfler
Member since:
2014-06-02

What about nuke codes, should they be handled by humans?

Reply Score: 1

RE: Comment by schrepfler
by modmans2ndcoming on Mon 23rd Jan 2017 00:31 UTC in reply to "Comment by schrepfler"
modmans2ndcoming Member since:
2005-11-09

Non sequitur

Reply Score: 2

Tesla should...
by lighans on Mon 23rd Jan 2017 09:44 UTC
lighans
Member since:
2006-01-14

Well, I think I tried to make a point. Sorry if it looks like trolling. ;)

I like biking. But I also like a great body (ahum). The combination is great.


I like to compare Tesla with Apple computers. The have a great design, with a talented salesman as CEO. But in the end it is just a car. Tesla didn't invent nothing, except popularised the idea of riding an electrical car. Making a car driverless is a sidestep in the process of making things AI.

But what Tesla should really do is making software that would increase safety of infrastructure and make it more friendly for humans instead of cars. But that wouldn't generate big bugs off course...

I would prefer a Tesla over a traditional car, though.

Reply Score: 1