Will driverless cars be good for cyclists or result in big restrictions?
 

When I think about driverless cars I'm confused about what to expect. Maybe having a car that can see and avoid cyclists will be the greatest thing since sliced bread. Finally, you don't worry about the texting/distracted driver taking you out. And pedestrians could cross the road without a care, knowing the cars will immediately react to their presence and avoid them. Right?

Well it's not likely to work out as nice as all that. Today part of your worry when you cross the street is wondering if drivers will see you and react in time to avoid you. So you tend to give the cars a wide berth and wait your turn for access to the road when you see a good time to cross. If you didn't have to worry at all, more people would tend to jaywalk at will, and that would bring the car traffic to a near stop.

When cars can fly around town without traffic lights what will that mean to traveling as a cyclist or a pedestrian?

See: http://www.bikeforums.net/advocacy-s...l#post17215765

Thanks - I should have searched and found that...

The car will slow down, frustrating the passenger, who can now give all their attention to throw insults/stuff at you.

My cncern is that driverless systems won't be used only to promote safety, which is a good thing. I think their main purpose is to cram more cars into crowded roads by enabling cars to travel both faster and closer together. Then there will be pressure to remove bikes from the roads.

I don't really think that's the purpose per se. I think at least a valid purpose is increased safety. But the net result is the same because the technology enables more crowded streets and really benefits overall speed by getting that.

I too worry about the long term. But unless somebody comes up with brilliant ideas that solve the energy crisis, I think bicycle riding is likely to only gain respect and hopefully we can demand and get more bicycle infrastructure as part of most plans that would impose limits on certain roads.

If this pans out it could do just that - but there's still some cause for skepticism.
Lockheed Claims Breakthrough on Fusion Energy - Scientific American

Cram more cars togther, that would be great. Instead of 2-4 lane roads we could get by with1-2 lane roads...that would leave room for dedicated MUPs in the other lane! More walkable and ridable cities!

That's an optimistic prediction. I hope you're right, but I think more cars is the likely result of higher road capacities. At least that's what always has happened previously.

I'm not holding my breath.

I'm pretty optomistic about what autonomous cars can and will do for road safety. Currently, two big obstacles are money and liability. GM doesn't wanna get sued by me because I got hit by one of their driverless cars, no matter who was inside. And if a driverless car commits a traffic offense, does the owner get a ticket? Could the manufacturer be sued for that?

I would actually expect less road rage as people engross themselves in something else and pay no heed to what the car is doing.

A little dose of reality to wake up the day dreamers:

"Perhaps one day tech enthusiasts will be able to visit a Museum of the Future That Never Was, where the Jetsons’ hover car and the Google super-robocar will sit side-by-side as showcase exhibits. Expect long lines for both, because the demos will be sensational."

Google self-driving car: It may never actually happen.

Yeah, I'm a software developer. If anything, the google car should be quite entertaining and/or frustrating (depending on whether it is your car) for quite some time at least.

It will be interesting when one is involved in an accident. Pranksters may even run them off of the road. What if you accidentally run into one? Will it wait around for a cop to make an accident report?

You beat me to it! I read that article and was going to post the link.

The main point was that Google's self-driving car relies on computer maps that are so detailed that they include every stationary feature along the car's route, including literally every lamppost. So far, only a few thousand miles of the millions of miles of US roadway have been mapped. The mapping process is both computer- and labor-intensive. It doesn't appear that Google's mapping hardware and software can create the original map databases, let alone constantly update the databases as real world features change.

I'm far from being a tech whiz, so I can't evaluate whether to believe Google's claims about the imminence of self-driving cars, or to believe this article.

From the above article:


I wasn't born yesterday.

I was somewhat surprised that Google set out with the initial goal of a fully autonomous car rather than having an intermediate step which strikes me as far simpler - full 'cruise-control' on freeways. I.e. you'd drive normally until you get to the freeway and then activate the enhanced cruise-control which would take over control until you neared your programmed destination exit. You could then nap, play computer games, etc. until a signal goes off near the exit and you resume control of the car. If the driver failed to retake control there'd be a planned safe-stop procedure where the car comes to a stop in a parking area and an alarm is generated so someone would come out to check on the health of the driver.

I agree.

I no longer believe that Google believes they are going to have a car ready for use by the general public anytime in the next decade. I think it's just a marketing ploy to beef up their image as the be all and end all of software development. If I had to follow all of the routes that their google maps and other software makes to the letter then I would not be sitting here. I would be stuck in the woods where google said to "go straight ahead" but there hasn't been a road there for 50 years.

This is just my gut-feel about it. I've been in software development for 35 years. There's hardly an environment more unpredictable than what we face on the road with many other vehicles and unique and unexpected obstructions that suddenly popped up that day. People can deal with this kind of thing with hardly a second thought. Software must have very specific, entirely unambiguous, and effective instructions that anticipates such issues. The problem can't be solved with GPS and a database of every object on the road. For one thing you don't get good GPS coverage in lots of places. For another thing such a database could never be complete enough to have the very latest data in it related to what happened today or yesterday.

And if you don't do it with GPS and a database then you're talking about software that dynamically processes video and recognizes the objects in the video and correlates that with sonar data, etc. and does all that before your car crashes into anything.

Not in this decade.

Good article:

I like the fact the Google car can't avoid a pot hole or ride in the rain or snow. That's pretty worthless if you ask me. Did you see the part that everything is going to be fine once they get the liability issue worked out? In other words, they want the same liability the motorist has when they cause an accident. No fault.

+1

You beat me to it! LOL

I love these visions of the future... even though none of them ever come true. I am often reminded of the huge lengths cities were going through in preparation for the increase in the number of horses. Then someone invented an affordable [horseless] cart that was powered by [something like] the newly developed lamp oil.

In 19th century France... this was a future vision of the rural postman in the year 2000.

http://bikeforums.net/attachment.php...hmentid=413677

That reminds me of elementary social studies class in the 3rd grade (1968). The text book showed a city in the year 2000. It included a shot where people traveled around the city in little vacuum tubes - a bigger looking version of the tubes they use at drive-thru bank tellers to send papers back and forth.

To see what people predicted makes them seem almost childish. You think "what were they thinking!". That's a good reason to apply some skepticism to almost everything we hear from "the experts" about what kinds of technology the future will bring - even the near future.

OTOH people invent things that nobody predicted. The internet comes to mind. We still don't have pervasive video phones like you saw in the 2001 movie. But we all have convenient access to a huge network of information and can interact collaboratively with the internet in ways that Arthur C. Clark and the like never dreamed of. Read old science fiction books. Nobody in those books was using anything like the internet. Sometimes people could submit questions to big all-knowing computers, like in Star Trek. But even then, it was not a social tool at all. And to do so you had to go to where that big computer or some kind of terminal was. People weren't walking around with handheld computers like our smart phones today.

The iphones Facetime is pretty much a video phone. My wife uses it all the time to chat with the grandkids. Although she'll answer with her ipad for a larger view when at home. I had bought one of the earliest (D-cam) Internet cameras and the conversion hardware to video chat from home using a 32" TV... back in 1995. It was a big expense (I even had to rent server time) with sketchy results.

But surprisingly.... most people prefer to text.... who would have ever predicted that. But connecting to our home security cameras through our mobile phones does seem to be Home Automation gold... it is extremely popular as a feature.

My Home Automation is designed around the Star Trek voice recognition model. Allowing me to turn on or dim lights by telling the "computer" to do so. I use an old XP laptop to interface with the hardware. Following the same ST model I get warnings and alerts via a female computer voice as well. I've used this set-up to retrieve Internet based information (like weather forecasts). But found I prefer a more graphic interface for that.

The most confusing part of the self-driving car [for commuting].... is what would be the purpose?!?! With the completely wired society why and where would there be a need to sit in a designated "work place" like people were restricted to in the last century. Even short range plans for public schools involve on-line learning... instead of old fashioned school buildings. And sending the kids to school in the self-driving car is one of their big selling points.

I think the self driving car.... solves many problems from the last century... that no longer need to be solved. The Johnny Cab? That may be a different story.

http://bikeforums.net/attachment.php...hmentid=413733

lol

Catch-22, but the mapping problem seems somewhat solvable by having many Google-cars rolling and collecting/uploading mapping data as they go. Sure, someone's going to be first on the scene of a change, but perhaps that can be solved as processing power and detection systems continue marching forward.

Also, just as people nay-sayed electric cars just a few years ago because of their limited range and long down-times, they are selling fairly briskly in spite of these continuing issues. Autonomous cars don't need to be able to go everywhere. How many people drive, or even care to be able to drive, all the roads in existence? If the early versions are restricted to even something like California and Nevada exclusive of the rural portions they would have a customer base in the millions, more than enough to get the ball irreversibly rolling.