Jump to content

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


Mallette

Autonomous Vehicles: Good or Bad  

49 members have voted

  1. 1. Are autonomous vehicles a good witch, or a bad witch?

    • Good
      20
    • Bad
      28


Recommended Posts

Mike, I certainly see no reason to change my attitude towards the value of human life just because others don't care.

 

That's the thing....I too very much do care.

 

If human life is valuable, then we wouldn't be requiring airbags. Equip every car with a roll cage and six point harness. Race car drivers are walking away from 150mph impacts with a concrete wall. Anecdotal in the delivery of my point, but every bit defendable...Air bags are stupid, but the propaganda has already prevailed.

 

Anyways, that's not my point.

 

My point is that I expect the human life risk to be higher with AV's without requiring lower speed limits. Yes, I know, I'm disagreeing with "the research" - but they're trying to sell a product. Nowhere in that research do you see analysis of the AV limitations and failure modes. That whole topic is just ignored by your "experts". I guess to be fair, it's hard to do failure analysis on systems that don't yet exist :P

 

Have you thought about what happens when the sensors start getting intermittent, wear out, break, environmental conditions, etc? At the end of the day, I simply won't be able to afford a safe enough AV. The economics of the automotive industry simply don't allow for it.

 

 

Btw, I would posit that my enjoyment of driving has very little to do with it being a high risk environment. In fact, autocross is probably one of the safest hobbies to enjoy. The enjoyment comes from an appreciation of the little details, and the gracefulness and finesse required to bring a car to the limit. It's very much like the audiophile pursuit of music. I'm not an SPL junky, nor a bass junky, nor a cold-hearted documentation junky. There's a finesse to finding the best racing line, just like there is a finesse to finding the best sounding system. It may require a lot of SPL and low frequency extension depending on the source material, but I don't want to give up the delicate nuances in the process. Racing may require a lot of horsepower and violent turns, but in the end it's really the fine touch on the steering wheel and pedals that I enjoy. I certainly don't subscribe to the big horsepower and sliding outta control all over the place - that's just over-compensation for a lack of true masculinity.

Edited by DrWho
Link to comment
Share on other sites

No problem with auto sports, but I am forced to play on Houston freeways 7 days a week and I am tired of that game. 

 

So, that car that completed the San Francisco to NYC trip did not exist?  Nor the 1.7 million miles Google claims?  And the car I rode in last weekend with lane control was all fake? 

 

So was the moon landing, I am sure.

 

As to what they are trying to sell, my belief is they'd like nothing better than if it really didn't exist.  But it does so they have no choice be to compete and accept a massive shift of liability they'd rather do without.  That is, if they are sane.

 

My personal belief is, I think, based on sound business.  They KNOW it works.  They KNOW people will insist on it.  They KNOW that if they dilly dally they'll be out of business.  They have no choice and they probably hate Google to pieces for forcing their hands.

 

Dave

Link to comment
Share on other sites

Dave, it might be cheaper, safer and faster for you to hire a chauffeur.

 

Just another idiot at the controls.  Why pay for it?  And where can I get a chauffer ready 24/7 good for the life of the car for 5k?

 

Dave

Link to comment
Share on other sites

I know of precisely one human I fully trust behind the wheel, my father in law.  Trained on 700hp cars at Daytona, and driven more million miles that I can count testing automobiles under extreme conditions.  He's in his 70s and they keep raising his pay to keep him on board because he's considered the best.

 

He says autonomous is better than him and even the simple automatic parallel parking is superior to his own.  He says that every time so far he's started to override AV controls it's already in progress before he can act. 

 

He has seen failures, largely minor, but that is because the stuff he's testing is often a couple of years from release...that's why he tests it. 

 

Between the common sense that a computer can make a decision much faster than a human and does so without road rage being a factor, I'll go with it. 

 

Dave

Edited by Mallette
Link to comment
Share on other sites

You aren't making points that mean anything to me

 

I think Mike's made a number of key contributions, but you're so fixated on your idea for the ideal mode of "safe-travel", that you can't see the forest through the trees.

 

Personally, I don't look forward driving in your area, and it's only for the money that I travel to Houston. The entire time I'm there I can't wait to get back home to small town USA.

 

 

I am glad you enjoy driving and are thrilled with high risk environment.

 

What risk is there in killing a bunch of orange highway cones?

 

 

As for me and my family I want a 90 percent better chance of not being a number on one of these Houston death total signs.

 

Did one of you're autonomous car articles actually offer a hard number with regards to a safety guarantee, predicted or otherwise?

 

I'd be interested in knowing how they calculated it?

Link to comment
Share on other sites

He says that every time so far he's started to override AV controls it's already in progress before he can act. 

 

What do you mean by this anecdotal example? All the AV overrides I've seen in vehicles take seconds to turn off....

 

Yes, TCS and ABS help even the best racecar drivers...but the dominant variable affecting the behavior of those systems is the human input itself. I don't put "driver aids" in the same category as "autonomous" - nowhere even close. Do you not see the fundamental difference?

 

Sit down and start writing out algorithms and it becomes very apparent. One accepts the user input as the desired action. The other needs to define the desired action. The hard part is that the desired action is always changing...it's how fast that definition can change that matters....not how well a defined action is implemented.

 

 

 

He has seen failures, largely minor, but that is because the stuff he's testing is often a couple of years from release...that's why he tests it.

 

 

I can't speak for every engineer out there, but usually when I'm testing prototypes, two things are true:

1) My design isn't finished yet. At the very least I've not yet proven the design meets performance targets.

2) The prototypes I use for testing are representative of the design intent. I'm not using prototypes that have gone through environmental extreme testing, etc...I'm testing with something that 'should be' fully functional....especially with early testing.

 

I think the main difference is our definition of "done".....if I could get paid to come up with ideas and stop at proof of concept, then my job would be very easy. Actually, I may be out of a job because that portion of the design cycle is maybe 1% of a design-engineer's total workload.

 

 

I'm still waiting on an explanation for how to deal with bad sensors. The failure rates today are in the range of 10ppm:

https://books.google.com/books?id=nlxbBAAAQBAJ&pg=PA157&lpg=PA157&dq=automotive+sensor+failure+rate&source=bl&ots=5eK0fHK2Wo&sig=JtDS1g9F_-NRWhWE204DHI6ajys&hl=en&sa=X&ei=-5RkVaSvJcGRyATs8YGYBg&ved=0CDoQ6AEwBA#v=onepage&q=automotive%20sensor%20failure%20rate&f=false

 

That's 10 out of every million cars that could suddenly no longer be able to make the right decision about the "desired action". If the failure is catastrophic, then it may be easy to detect, but what about all the brown-out or intermittent problems? Call that 1ppm....many of those will be early in the life of the vehicle, and then the rest is just a matter of time until age takes its toll. I think it will actually be much higher because as other reliability research indicates, it is measured by that which is detected by the end users. I think modern failure rates are much higher, but it gets masked for a variety of reasons.

 

The problem is that this is the part of the world that only the engineers really get to see....

 

What's going to happen is the AV world is going to continue their safety propaganda....eventually there will be massive deaths, and then a lot of finger pointing, and then maybe a few laws written to "prevent it from happening again" and the cycle will continue....and as long as we keep the deaths from drunk driving, texting, and other forms of distracted driving high, then we'll have numbers to compare against that make the AV look good. In the meantime, the straightforward solutions to drunk driving / texting / distracted driving will be completely overlooked because there is more money to be made in the autonomous world. The automotive death rates are really quite low once those three are solved....

 

The reality is people want to be able to text and drink while they're being transported from A to B....and to that extent it is totally a convenience argument.

 

 

How about airbag failure rates?

http://www-nrd.nhtsa.dot.gov/pdf/esv/esv21/09-0199.pdf

1-2% should have deployed scenarios resulting in fatality....what is that, 10,000 ppm? Hmmm....and I would have thought airbag deployment would be a much simpler decision to make....

 

 

Should we get into the MTBF (meantime before failure) calculations that exist for all electrical and mechanical components? We as engineers actually know things will eventually break, and we can predict it with fairly good accuracy too. The fun thing about MTBF is it is a bell-shape curve....some things last longer "than they should" and some things break way earlier. What should that number be? At what point do we accept a shorter number because it costs less and helps make our product more competitive in the market place?

Edited by DrWho
Link to comment
Share on other sites

Mike, your objections simply don't translate to the real world.  Odd that it seems I am looked at by you as a starry eyed optimist or something but what's happening in the market as well as in testing all supports the relatively conservative prediction I made. 

 

Heck, Gilbert wants to know how the 90% reduction in deaths was arrived at and by whom.  Well, seems to be pretty much all involved as it is pretty uniformly applied.  I think it's because to those of us who believe in technology it appears extremely conservative.  I'll let you guys dig up the scoop on that. 

 

Point is I really got sucked into a defense of an emergent technology.  All I started out to do was to present some interesting technology news.  I was completely taken aback by the lashing out of people in love with their bleeding cars.  I really love my gas range in the kitchen...but it's hardly a piece of my being.  Just a tool.  Tools are good, but they are just tools and replaceable with better tools. 

 

One person here almost violently stated their devotion to a vehicle I would not be caught dead in, and driving such a car in Houston would be a great way to be caught dead in it. 

 

As to air bags, I will keep mine, thank you very much.  The failure rate of air bags seems to be much lower than that of craniums. 

 

I am going to continue to post here as new things happen, but pay less attention to the anti-progress crowd. 

 

Dave

Link to comment
Share on other sites

Just a tool.

 

There-in lies the difference between us, and all the people you're call "anti-progress", and "not caring about saving lives". With comebacks like that, you should have anticipated an assault-like verbal back lash.

 

Like many people, I love my vehicles, even my bicycle, my roller blades, they are all tools to be enjoyed, tools for relaxation, tools that take my mind off my work. All my vehicles serve multiple purposes, all of which include transporting me from one place to another, but they are so much more than just another tool. 

 

I would never compare any of my vehicles to my kitchen stovetop. I feel bad for you, and I know your in traffic hell.

 

If I ever felt that way, then I would surely know it's time to relocate, or find another mode of transportation. I hope you live to see, own and drive (you'll have too) what you think is going to be your dream car. But I honestly don't think it take too long before you're complaining about it's short comings.... which will be significant, relative to all the comments you've made of what you think it will be like.

Edited by Gilbert
Link to comment
Share on other sites

There-in lies the difference between us, and all the people you're call "anti-progress", and "not caring about saving lives". With comebacks like that, you should have anticipated an assault-like verbal back lash.

 

Gilbert, pleased to apologize.  Just did not understand the connection people have to their horses or whatever.  Still don't.  But I respect your attachment and lack of understanding doesn't mean a lack of respect for your opinion.

 

As I said, I intend to largely refrain from posting here other than news of the technology.  Given what is already out there and the announcement from people who have no reason to embellish I certainly am more optimistic than you are. 

 

Anyway, the proof is in the having, and next vehicle I intend to have as close to state of the art as I can afford.  My belief is that it will be at the level of the car that just completed the coast to coast about 98 percent hands free given that it will be a minimum of two years or more before I buy again. 

 

Dave

Link to comment
Share on other sites

Oh please, no apology necessary. not with me, and certainly not about this subject. I understand your opinion, I understand the shyt you have to drive in every day, and I understand your desire to just want to sit back and let someone else do the driving. But I'm sorry to say, that I'm pretty darn confident the autonomous car is not going to be the cure you think it is.... hopefully we'll live to see some on the road, but I won't be one of the people buying one, when I'm 90 yrs. old, maybe, but doubtful.

Link to comment
Share on other sites

Mike, your objections simply don't translate to the real world.

 

Last I checked my data came from the real world???

 

 

 

I really love my gas range in the kitchen...but it's hardly a piece of my being.  Just a tool.  Tools are good, but they are just tools and replaceable with better tools.

 

 

For car guys, the cars they enjoy are part of their identity....it's not seen as just a tool. Same goes for pretty much any tool out there...just look at mechanics and their name brand association with their wrenches....

 

Btw, I'd argue you do the same with audio too (as do the rest of us).

 

 

 

I am going to continue to post here as new things happen, but pay less attention to the anti-progress crowd. 

I might suggest that I'm not anti-progress.....I'm just gunning for what I believe to be a better solution. Disagreeing on the path of progress is a lot different than being against progress itself.

 

 

Btw, I started researching the statistics about the causes of accidents and fatalities, but I'm having a hard time nailing down reliable statistics and really I should be getting back to work. But here's a few that are interesting from 2012:

 

Total deaths = 30,800

 

18,705 were single vehicle accidents. Of the remaining 12,095 multi-vehicle accidents, 2,902 were alcohol related.

(http://www-fars.nhtsa.dot.gov/Crashes/CrashesAlcohol.aspx)

That means there are at most 9,193 deaths per year caused by another driver that isn't alcohol related....and that number is probably closer to 5,500 based on other anecdotal references that I'll eventually parse together someday. Just to reiterate the point, about 1/6 of the automotive deaths were beyond the control of the victim.

 

This summary given to Congress several years back indicates some interesting percentages too:

http://www-nrd.nhtsa.dot.gov/Pubs/811059.PDF

Driver Fatigue and distractions are the dominant causes cited.

 

 

Putting it all together, I'd argue we could put a huge dent in 90% of the deaths caused by other people. The people killing themselves? Hey, I thought evolution was the religion of choice these days.....why not let natural selection do its thing? But seriously, if we're truly safety oriented, then why don't we take the low hanging fruit that could already be implemented today for a much lower price? Looking into it, it either comes down to greed or entertainment....and that's the same thing driving the AV stuff.

Edited by DrWho
Link to comment
Share on other sites

Anecdotal story....

 

A buddy of mine was towing a trailer the other day and just about destroyed his transmission because he kept it in overdrive...it was constantly running between gears and just roasted the torque converter. Long story short, in ended up in an interesting discussion about why people don't read their user manuals.

 

There seems to be a strong aversion these days to educating yourself on any subject. I seriously wonder what the root cause is, but I very much see the AV world as another step in the direction of avoiding education, and "just letting the system do it for me". I'm not sure all this newfangled technology is really pushing our culture into a better direction - well I guess better is a loaded term, but quality of life is not proportional to the quantity of technology. I find it quite ironic that as an electrical engineer I think society should move in a direction away from tech gadgets.

 

For example - the other night some friends were helping me remodel my house. We went out for a nice dinner afterwards and everyone was on their phone for the entirety of the meal. In fact, they were on their phones while working on the house too. Always connected to somewhere else other than the reality right in front of them. It's way worse with the younger generations who are checking their phones every 5 minutes like they have a nervous habit.....and we wonder why these people crash into stuff on the road. So we make AVs so that people can be on their phone while commuting....and for what, to play stupid games? read Facebook? Stay "connected" with their "friends"?

 

Maybe I just gotta find a new set of friends? The problem is the Amish people don't like cars - and thems a lot of fun to drive ;)

 

I dunno, I'm definitely biased by my larger cultural views, but I feel I gotta take a stand for every little bit of sanity left in this world. I really don't want to be a mindless creature roaming around with everything being done for me. It's kinda nice to actually do stuff with your own hands every once in a while, and driving is one of the last islands I can always retreat to.

Link to comment
Share on other sites

This just in on the self-parking Volvo (I'm pretty sure that guy got one of his knees taken out, but the article I read says no injuries were sustained..... the parking video is followed by a not so new Volvo self-braking video...... followed by a 3rd. video that's a creampuff of a sales-pitch for why you need to buy one.

 

http://www.bing.com/videos/search?q=volvo+autonomous+car+demonstration+goes+wrong&FORM=VIRE2#view=detail&mid=922B3AC838CF809B0C0A922B3AC838CF809B0C0A

Link to comment
Share on other sites

Oh look! You get to eat an apple in the car...haha.

I love how they said "faster than most drivers"....Apparently they're not willing to stand behind being faster than everyone. Suppose they're using Google Maps as their 3D cloud based map?

I'm really surprised by that car rear ending the trailer....it looked like perfect conditions all around. I wonder what the cause of failure was. It didn't even appear to brake.

Link to comment
Share on other sites

I think I said right off the bat there would be lots of stories and videos of failures.  Disruptive technology always has those hoping it will fail and pushing the occasional failure as though it were the rule.  In some cases, they even create them when desperate enough.  I remember all the horror stories in the early sixties when seat belts were mandated.  Sounds just like this stuff...only worse. 

 

Family burns to death because of seat belts! 

 

Young couple found still strapped in on bottom of lake!

 

Blah, blah, blah. 

 

Dave

Link to comment
Share on other sites

I think the better analogy would be a seal belt demo where the belt ripped in half, or the buckle let loose completely. Basically complete failure of proof of concept.

 

Crashing into a parked vehicle during a demo? You're seriously going to defend that technology? Nothing worked! Man, I'd hate to be the engineers at Volvo responsible for that screw up. I've no doubt it was something stupid simple, but that's the problem with complex systems....it's the stupid simple that cripples it.

Link to comment
Share on other sites

Yes, Mike.  We should have realized we didn't belong in space after Vanguard. 

 

I mean, if you can't get something right in one shot why bother.

 

As to complex systems, how's that German pilot for a complex system crippled by something stupid simple?  :P

 

 

Dave

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...