Autonomous Vehicles

Professional Engineer & PE Exam Forum

Help Support Professional Engineer & PE Exam Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
http://www.denverpost.com/2016/10/25/self-driving-beer-truck-colorado/

This is already happening. The first step is connected vehicles, which exist. Three transportation agencies have pilot projects for connected vehicles (http://www.its.dot.gov/pilots/). A car doesn't need to see a sign, but it will need pavement markings. Dow is already working on a marker that can go in paint, rather than an expensive transmitter under marking tape (like 3M will try to sell), that can be "seen" under snow. 

The old saying in traffic engineering is that the you can't make the road totally safe because of the nut behind the wheel. Autonomous and connected vehicles remove that nut. 


Connected vehicles work great when they're all connected, but we have nearly 250 million cars that aren't connected, so what do we do with those?

Markers in paints sounds like a promising idea, but we still have lots and lots of streets and roads without lane markings. . .

 
Okay, I suppose all the "experts" say everything would be better because, in the event of an accident, a computer would be the one to determine the best course of action to take, i.e. do you hit the car with one person in it or the van with 10 people in it.  Now obviously a computer is going to hit the car because it is 1 person as opposed to 10, but what if the car only has one person in it because it is loaded with puppies being taken to visit a nursing home while the 10 people in the van are white supremasists on their way to a klan meeting?  How smart is your computer now?

Also, it's bad enough getting the blue screen of death when I'm sitting at my desk at work, what's going to happen when you get on going down the highway at 70mph?

 
relying on paint doesn't sound like a good plan either.  The maintance schedule/cost of repainting would be crazy.  How much can get worn away before it becomes an issue.  small towns don't have the budget to repaint lines on a routine basis.  who will pay for it?  

 
This is why autonomous vehicles still have driving capabilities. There will be areas where connection is lost or paint is lost and the driver has to take over. 

The issue of connected vehicles is not an all or nothing. It's a bonus to vehicles that are connected, but older vehicles are not harmed. Also, there is definitely going to be the ability to add connectedness to existing vehicles. In fact, XM radio is partnering with the project, meaning as long as you have satellite radio, you can tune to those alerts. 

While engineering is a profession built on age old theories, it's also a profession that seeks to improve the world and think past what we know now. I'm really excited to be part of a project that is looking at the future. 

And, on the ethics front, Mercedes has programed it's car to protect the driver (http://blog.caranddriver.com/self-driving-mercedes-will-prioritize-occupant-safety-over-pedestrians/). Bold move? Yes. However I'd argue this is how most people are driving today. You don't take the moment to judge if you should kill yourself or kill the pedestrian. You naturally preserve your own life. 

 
that's the way I see it, as an added bonus to people that can afford a $80,000 Tesla.

But they way people talk, its like there wont be any more wrecks because all cars will lock in place digitally to drive to grandmas on the interstate (not going to happen any time in our lifetime).

My company is also "partnering" with the Feds to research some of these things, as I sat through an overview of our "work" all I could think of was "These people have never been through a 7 hour deposition over a fatality before"....

 
My father taught me to drive by making me drive entirely in reverse in an 85 Toyota pickup with a 5 speed with the worlds worst gearbox and stiffest clutch release imaginable.  He made me drive backwards laps around the nearby middle school, doing loops around islands, etc.  His justification was "if you can do this, then the rest of it will be easy."
I was taught on a 1970 Pontiac GTO. Arguably equal in difficulty with the gearbox and clutch stiffness. 

 
This is why autonomous vehicles still have driving capabilities. There will be areas where connection is lost or paint is lost and the driver has to take over. 

The issue of connected vehicles is not an all or nothing. It's a bonus to vehicles that are connected, but older vehicles are not harmed. Also, there is definitely going to be the ability to add connectedness to existing vehicles. In fact, XM radio is partnering with the project, meaning as long as you have satellite radio, you can tune to those alerts. 

While engineering is a profession built on age old theories, it's also a profession that seeks to improve the world and think past what we know now. I'm really excited to be part of a project that is looking at the future. 

And, on the ethics front, Mercedes has programed it's car to protect the driver (http://blog.caranddriver.com/self-driving-mercedes-will-prioritize-occupant-safety-over-pedestrians/). Bold move? Yes. However I'd argue this is how most people are driving today. You don't take the moment to judge if you should kill yourself or kill the pedestrian. You naturally preserve your own life. 


I've never said it's not a good thing to work on, but I don't see it being quite as revolutionary as some are trying to make it out to be, either.  At least not in the next 40 years or so.  There's a huge numbers of bridges we need to cross (pun intended) before we get to the grand vision of the car doing all the driving for us.  And it is a technology that IS going to kill people, so there's a lot of thought and consideration that needs to go into that.

And no surprise on the Mercedes program.  If you're driving a Mercedes of course you think the world revolves around you!  I can already hear the lawsuits about how this discriminates against lower-income demographics. <this is somewhat tongue-in-cheek>

Curious if you feel comfortable riding your bicycle on the road with self-driving cars?

 
Curious if you feel comfortable riding your bicycle on the road with self-driving cars?
I have no clue. Maybe! The drivers in SF are in general not the smartest, and drive very stupidly and unpredictably. A computer might do better.

 
Last edited by a moderator:
Yes, I think I would. 

Here's the kicker- an autonomous vehicle doesn't "feel" anything, as opposed to drivers that are angry I'm on my bicycle. Riding home on Saturday night I had a car speed up and swerve towards me before turning at the last minute. They were obviously trying to scare me and it worked. An autonomous vehicle has no reason to taunt me and it will work to avoid me. An autonomous vehicle removes a lot of the variables. 

 
That goes both ways, though.  The autonomous car will have no feelings as it runs you over because it didn't "see" you, or was not programmed correctly to avoid you.  Whereas I believe fundamentally very few drivers would deliberately try to run you over, even if they don't like you or are angry (and yes I've experienced some of these incidents as well).  Or someone who really doesn't like bicycles could hack a car and make it deliberately target you, or introduce a virus that does the same.  Given my own exposure to AI and programming (granted I'm no expert by any means), I wouldn't trust my life to this technology.  I'm just a crotchety old engineer though.  Get off my lawn.

 
I had an introductory course that touched briefly on AI. It's both fascinating and terrifying at the same time.

 
With the amount of distracted drivers out there nowadays, I'm wondering what the failure rate of a driver not seeing you is in comparison to a vehicle not seeing you. 

In either case, it will still be the cyclist's fault. 

 
I'll tell you what.  We just bought a vehicle with lane keeping assist.  Now, I don't know about you, but I move toward far line when some vehicle is crowding the line next to me.  But of course, the LKA system doesn't think that's an appropriate move.  Lane centered at all times!!

I will agree that a certain percentage of drivers' driving would be vastly improved by a car that drives for them.  But there is a host of drivers that wouldn't see any benefit (save for perhaps NOT having to be attentive the entire trip) over their driving skill.  I'm not sure it quite balances out.

 
My old Ford Explorer had that automatic cruise control.  It was awful.  In theory it's a nice idea, set your cruise control and if someone in front of you is going a little slower, the Explorer would slow down without having to tap the brakes and cancel the cruise control.  In reality, apparently Ford decide that about 1/4 mile was the correct distance between cars.  Sure you didn't have to adjust your speed because of the slower car, but in the meantime, cars would keep cutting in front of you because of the giant gap.  Guess what...every time a car got in front of you, the Explorer would slow down even more to maintain that ridiculous gap.  Even if there was nobody getting in front of you, you were so far back that most of the time the car in front of you would never move cause they figured you must be going the same speed and you'd find yourself cruising along 5mph slower than what you set the cruise at.

 
^^ I know some vehicles allow you to have some flexibility with how "aggressively" the auto cruise control follows the car in front.  Allowing the driver to select how closely the car follows would cut down on that

I don't know how soon, or whether cars are ever going to be totally autonomous but I am impressed with how quickly a lot of these driver aids are being implemented.  Automatic braking, blind spot alerts, auto cruise control, lane departure warnings, etc. are all becoming more and more common on cars today.   And given how many people I see texting while driving i am all for more and more automatic safety equipment on cars.  

 
Last edited by a moderator:
Back
Top