Poll & Prediction: Autonomous Car Equipment at 5k by 2019

[Khornukopia]

6 minutes ago, Mallette said:

Same as with any other obstacle or issue.  They will back off to their own safe response distance, which is closer than a humans as their response time is much faster.

 

Dave

But if they back off every time a human driven car causes an unsafe following distance, I think there may be times when the self driving cars will slow down so much that they may just stop in the middle of the heavily travelled freeway, because they can't respond to the chaos of human drivers. Not trying to be negative about self driving cars, which I find fascinating, I just see a lot of commuters who think they are NASCAR drivers, with a phone in their hand.

[Mallette]

5 hours ago, Khornukopia said:

because they can't respond to the chaos of human drivers.

Inner lanes will be restricted to AVs pretty quick after they start to go mainstream.  Beyond that, AVs will also change lanes if there is a better speed option available.  What do humans do in the same situation?  Get mad, brake check, and in Texas even pull out a pistol and shoot the perp.  Personally, I think the AVs will handle it better.

 

Dave

[vasubandu]

9 hours ago, Khornukopia said:

I just see a lot of commuters who think they are NASCAR drivers, with a phone in their hand.

 

That is why I think that that transition will be very short until people-driven cars are outlawed or limited to Antarctica.  Mixing cars would be like a bank having its employees do some balance calculations because the machines cannot be trusted.  People like to point out that space shuttle cannot land on autopilot, but is has computers less powerful than your smart phone and that had be designed for the rigors of space travel, so it is not fair point. The question is why would we let people drive cars when AVs are around? It is an activity dangerous to others that will be totally unnecessary.

 

The other thing is that we and China will be able to make all the AVs we need very quickly.  They are likely to be 20-25% of a small car.  They can be made cheaply in the same respct that computers can.

[Mallette]

8 hours ago, vasubandu said:

That is why I think that that transition will be very short until people-driven cars are outlawed or limited to Antarctica. 

I tend to agree, logically.  The curve of perfecting hardware and software will be pretty much the fastest ever due to the massive amount of data as the numbers grow.  There will be accidents, as we've now had a handful, but every one will feed data back and issues eliminated.  It took well over a century to improve our highway system and its controls, and vehicle safety, to the point it is.  This will be FAR faster as the OEMs, given the liabilities of hardware/software bugs, will have more impetus to improve systems than ever before.  In the past, OEMs have dragged their feet on many things from seat belts to efficiency, but this will be different as their very existence depends on it.  Some airlines went under and others struggled after accidents in the 40s and 50s, Lockheed pretty well lost its airliner trade after the Electra crashes in the 50s.  Automakers must get this right or they will die trying.  

Dave

[CECAA850]

Sure Dave this will be different.  As long as bean counters are still running the major auto makers it'll be same ol same ol.

[Khornukopia]

The upcoming trial of Google's Waymo vs. Uber, about their self driving car LIDAR technology, should be interesting.

[Mallette]

33 minutes ago, CECAA850 said:

As long as bean counters are still running the major auto makers it'll be same ol same ol.

Yes, it will.  That is what I am counting on.  

1. Bean counters are paid to keep costs down.

2. As demand increases for AV abilities, OEMs must respond or die.

3. Bean counters must respond by ensuring that the vehicles are as liability proof as possible.

Like Walmart, Sears, and others, it's either respond to the market or die for automakers.  Most will respond, some will die.

 

Dave

[Mallette]

1 minute ago, Khornukopia said:

The upcoming trial of Google's Waymo vs. Uber, about their self driving car LIDAR technology, should be interesting.

We badly are in need of a cost breakthrough in LIDAR.  While alternatives are working well, LIDAR is by far the best option.  I am sure millions in R&D are being spent on it.

[CECAA850]

2 hours ago, Mallette said:

Yes, it will.  That is what I am counting on.  

1. Bean counters are paid to keep costs down.

2. As demand increases for AV abilities, OEMs must respond or die.

3. Bean counters must respond by ensuring that the vehicles are as liability proof as possible.

Like Walmart, Sears, and others, it's either respond to the market or die for automakers.  Most will respond, some will die.

 

Dave

Items 1 and 3 are what's currently happening yet all auto manufacturers are issuing recalls at break neck pace.  It costs them millions.  The reason for the recalls is your point number 1.  Parts supplied by the lowest bidder and not enough R and D.  Why?  See point number 1.  The stakes will be much higher when cars drive themselves yet bean counters will still be in charge.

[Mallette]

I understand, Carl.  But you are overlooking the fundamental economic role of demand.  People like me are demanding autonomous functionality and see no reason to purchase a vehicle at this time that doesn't offer these safety margins.  What I have is fine until I feel like there is something compelling to purchase, like a car that has the ability to keep me alive when I don't.  A decade ago, side air bags and 5 star crash ratings were the domain of the well to do.  Today, my decidedly bottom of the line Elantra has both as well as a 44mpg efficiency on the highway.  What reason do I have to purchase a car that won't improve on that?
 People are not vegetables and they don't care what the bean counters or low bidders say.  They purchase on price/performance and many consider survivability to be part of that equation.  

 

Dave

[CECAA850]

I feel people like you are in the minority. 

[Mallette]

Again, you aren't getting it.  We ALWAYS start in the minority to start.  I was an early adopter of CDs.  Few would purchase a car with airbags at first, or seatbelts.  But now nobody would purchase a car for their family and kids without them.  Few today understand the advantages of just forward looking brake assist and lane control, but as they realize it makes them many times safer they'll catch on.  It's how the market works.  "...people like me.." are those who are not dreamers, but see the advantages these systems have in preventing accidents.  Federal, state, and private entities like insurance companies are projecting continuing and accelerating reductions in highway deaths and accidents, and it certainly isn't due to us suddenly developing higher skills or better reflexes.  And the rate will continue to accelerate as people catch on.

 

Dave

[CECAA850]

When did I ever say it was inevitable? 

[CECAA850]

I just don't have on rose colored glasses.

[Mallette]

12 minutes ago, CECAA850 said:

I just don't have on rose colored glasses.

Some sort of tint, as you are not thinking about established economic principles.  I cannot say I am entirely objective as that would prove I am not.  I can say I make every effort to look at tech and other trends through the light of history.  What I've heard of repeated history here are the arguments of the horse enthusiasts of the early 20th century.  Certain a huge majority to start, but only a fading memory now has history moved relentlessly to ever higher technology.  Whatever glasses you are wearing, try to take them off and look at the facts.  

 

Dave

[Mallette]

This is the first "evergreen" thread for me.  Hardly with the "big ones," but I get the feeling it may go on for years with the paddles being put to it from time to time as things develop.  This IS a potential paradigm shift potentially even greater than the automobile itself and watching it will be fascinating.  

Dave

[Mallette]

Gonna make it easier to purchase great Klipsch speakers from far away.  Just send an AV to collect and deliver them.  

Dave

[Arrow#422]

6 hours ago, CECAA850 said:

I just don't have on rose colored glasses.

I have a pair you can borrow, but beware because the vision they provide is very "rosy"  and not necessarily accurate of the immediate surroundings.

[CECAA850]

16 hours ago, Mallette said:

Whatever glasses you are wearing, try to take them off and look at the facts.  

I look at the facts and relate what I see every day.

[Don Richard]

The facts are that companies are working on AV solutions and have made progress. However, there is much more work to be done on the sensor arrays and particularly the software that interprets the sensor inputs and matches the sensor patterns with real world traffic scenarios. This work is being done at test tracks around the country, test tracks that have urban, suburban, and rural sections. Level 3 AVs are now being tested on public roads. Data has been gathered and will be analyzed and studied. Control software, possibly integrated with AI, is being written, refined and tested. As always, safety is a prime concern so the software will need redundancy and fail-safes, and be certified before Level 4 automated vehicles will be released for public sale. The probable timeline should be 3-5 years for Level 3 AVs ready to be sold, about 7-10 years for Level 4. If manufacturers rush AVs to the showrooms before they are ready and accidents occur,  these times will doubtless increase.