Boston to start planning for driverless cars
Even as we still try to tame our human drivers, Boston will begin planning for cars that drive themselves - in a year-long effort that will include figuring out how to test "autonomous" cars on our centuries-old roads.
Mayor Walsh this morning announced an initiative, with the help of the World Economic Forum and the Boston Consulting Group, to get Boston's roads ready for the cars of tomorrow.
Over the next year, the city, the consultant and companies developing driverless cars will try to develop both on-road tests and policies for dealing with the cars once they become widespread.
Nikolaus Lang, senior partner at the Boston Consulting Group, said in a statement:
Shared, autonomous vehicles have the potential to fundamentally improve urban transportation by enhancing accessibility for the city's residents and increasing road safety. We are excited to be engaging with the City of Boston during the coming months on making this vision for urban mobility a reality.
Ad:
Comments
Driverless cars
Will driverless cars be able to "give the finger" to other bad drivers? A Boston tradition that dates back to horse and buggy days.
Yep
And all the driverless cars will be equipped with a robot hand with an extendable middle finger with which to return the greeting.
More importantly can they double park?
There needs to be an override feature for this or this plan is a non starter.
Oh, and talk to me when there's a self-shoveling car. I'm in.
Backseat drivers
Remember, these cars will have passengers — intrepid souls with both hands free to express themselves. Except when they are assuming the crash position, that is.
So maybe not so much.
(I like to imagine driverless cars as completely empty, roaming around aimlessly, deflecting all the elementary forms of road rage.)
Ideally, someday most
Ideally, someday most driverless cars would be doing just that--they would be car shares, driving around and giving people rides instead of spending 90% of their time idle in a parking lot
Yeah,
drivers in other cities, states,countries are saints and never engage to behavior like giving the finger.
Where do you people who love to whine about how bad people from Boston and MA come from? Mayberry?
The point has already been well made...
But, yes, they will have a Masshole plugin.
And self driving cars can't possibly kill any more people that people driven cars already have.
Driverless bicycles
are what is needed along with preparation for traffic to go up 20% because fewer people will put off trips because they hate driving in heavy traffic. So, more vehicle lanes combined with longer turn queues and storage.
Eventually, with only intelligent or driverless cars and driverless bicycles, traffic lights wouldn't be much used, just vehicles telling each other when to go. It looks chaotic in simulation videos but that makes it safer for pedestrians and non-autonomous bicycles who will have to wait for their signal to go, otherwise face great danger in what really looks like Frogger. The autonomous vehicles will try to stop for Boston jaywalkers, but may not always succeed and result in gridlock.
Massachusetts
has the lowest auto and pedestrian fatality rate in the country.
People historically jaywalk a lot in Boston because of the particularly un-American-like, non-grid layout of the city. The density and congestion also plays a role. Humans adapt to, and shape, their environment.
What?
I say again, what??
Despite the highest crash rate
The crash rate is the highest - look it up. Shortest mean time between accidents for our drivers.
Test them in the North End
Test them in the North End and on the Jamaica Way.
Who Gets Killed First?
1. Bike Messenger going wrong way down Franklin Street.
2. Pedestrian when car ignores the no cars rule on Washington Street between Temple Place and Franklin Street. (The Jehovah Witness people are smart, they stay on the sidewalks there, they will live).
3. Passenger in the JohnnyCab when its CPU blows up because it can't understand the concept of weaving to avoid UPS and FedEx trucks bringing Boylston Street to one lane.
4. BU Student inside when the prototype fails to understand concept of other BU Students crossing streets, falling out of windows, through skylights, or stepping in front of commuter trains and expecting the other party involved to stop for them.
5. The car itself when it violates a bike lane and gets whacked with a kryptonite lock for having the audacity not to obey the selective use of road rules by area bike riders.
Good luck to these people. You are not going to like this but, these cars are actually needed more in the suburbs than the city owing to the city's better transportation options.
Probably the bike
I know your post was mostly in jest - but apparently these cars are very good at most driving skills. The biggest problems seem to be with bikes (because they are as small as a pedestrian but move as fast as urban traffic) and snowy conditions.
The Tesla that crashed in FL a few weeks ago was apparently confused because it couldn't distinguish between a flat, white sky and the white side panel of an 18-wheeler (purportedly).
Had an interesting discussion with someone yesterday about the "moral" implications of the software. For example - it will be sophisticated enough to determine that it's about to hit a pedestrian and it can't be avoided unless it involves an accident that will kill or severely injure the passengers. What choice does the car make? Does it value the lives in the car - for example, if the car has an 89 year old person in it and an "interface" with the pedestrian's cell phone tells the car that the pedestrian is a world renowned heart surgeon - does the software make a "moral" choice to swerve and hit a pole, or run down the doctor?
All sounds a little sci fi and quite frightening - but we are literally on the footsteps of that world and real world cars will have to make that decision probably within our lifetimes.
A Boston of self-drive cars
A Boston of self-drive cars would be a big attraction to world famous surgeons. They would not need to obey any traffic laws. For instance, they could jay walk at all times and all the self-drive cars, knowing who they were about to hit, would steer to crash into someone else.
Wherever they walked they would leave a trail of crashed cars and debris behind them.
Moral decisions
Here's a fun tool from MIT to play around with if you want to try this yourself:
Cyclists welcome this concept
With bunches of cars roaming around rather than have a place to park, and people abandoning public transit to hire these things, there will be completely gridlocked traffic all the time. That means that cyclists can go where they want and do what they want because nobody else will be able to move.
That isn't why the Tesla crashed
The crash occurred because the left-turning truck utterly and completely failed to yield the right of way.
So what?
This is a thing that happens. Even in Boston.
winter is coming
And the ultimate question is:
will the self driving cars be programmed to recognize Boston's written and unwritten space saver laws?
Interestingly, fixed gear
Interestingly, fixed gear bikes found a way to completely screw with a self driving Google car in Mountain View, CA. One was waiting behind someone track standing on a fixie, and the car kept getting confused as the cyclist rocked back and forth. Makes me wonder how these cars react to small in place movements of pedestrians as well.
I'm adding #6. The first
I'm adding #6. The first responderswho try to stop the car because they see a car rolling with no driver go past them.
6. The driverless car that
6. The driverless car that doesn't know to stay far, far behind that 13'-high truck in front of it on Storrow Drive.
Autonomous cars
Will they be able to do or recognize the "Massachusetts Left"?
Probably get rear ended a lot
From this NPR report from Pittsburgh it sounds like the cars are super cautious which will take a lot of Boston drivers by surprise.
http://www.npr.org/sections/alltechconsidered/2016/09/14/493823483/self-...
Anything that goes the actual
Anything that goes the actual speed limit will get rear ended.
It's still a long way off,
It's still a long way off, but I eagerly await the day when a person driving a car seems as foreign a concept as a career as an elevator operator.
pay attention Uber/Lyft drivers
Uber drivers who took on car loans need to be concerned. Uber drivers who are locked into a vehicle subprime loan need to pay even closer attention. As soon as Uber can cut out the driver and maximize profits they will. Anyone barely scraping by is going to get left in the dust, as is usual, while the corporation makes a healthy profit.
Yep
That's why I didn't borrow money that I don't have to acquire a better vehicle when I gave Uber a try.
Two fare cuts and an unwillingness to take Pool pings later, I was gone, but I still have a slightly nicer vehicle that I can pay off today if I want to.
actually
read an article recently speculating that the Auto industry will probably offer their own ride sharing services once they have got these driver less cars. That will cut out the middle man (uber) and will sell these cars after 3 years, like leased cars. Win win for them.
Pretty much what Car2Go is: a
Pretty much what Car2Go is: a repurposing of Smart Cars that were not selling fast enough. You still have to drive them, but a pretty cool idea which gave ZipCar a run for its money in a bunch of cities.
Ah nice
Cars that follow the speed limit and stop at red lights
In other words, cars that
In other words, cars that will be rear-ended constantly.
Jaywalkers will game this!
Why stop walking when the light changes when you have a large group of people and the driverless cars will be the ones first at the light because they actually stop and don't block the box?
Flocks of pedestrians at rush hour will just lock down the city by keeping the driverless cars from moving.
Looking into my crystal ball
This is how it will play out. Someday in the next 10 years, a kid will run into the street and get killed by a self-driving car. Incredible outrage will ensue. Legislation will quickly be passed putting onerous restrictions on the details of how and when autonomous vehicles can operate. As a result, only highway use of self-driving cars will be practical.
(As I have said many times to my wife, the day will never come when you hop into your self-driving car in Hyde Park have it drive you to North Station. Humans are too unpredictable and all the technology in the world cannot make up for the fact that stupid divers, bikers and pedestrians will always be too much of an impediment to making self-driving cars sufficiently safe.)
You don't learn from history.
You don't learn from history. When motor vehicles began to be used on public streets, there were deaths and ensuing outrage. Drivers were beaten and lynched by mobs after injuring people and automobile prohibition groups sprang up. But thanks to lobbying by oil, steel, and rubber industrial concerns and car makers, legislation against cars got nowhere. Eventually that lobbying yielded the car centric culture and everything that goes with it that we have now. Isn't it obvious that Tesla, Microsoft, Google, etc. are just the latest players in the ongoing battle and not necessarily of a separate faction at all ? To that list I would add government, which has been planning for and developing ubiquitous smart tolling for years, quietly and without secrecy. Automobiles have always been about separating people from their money, not about caring if somebody lives or dies. What's going to change that now?
Self-driving cars will end up
Self-driving cars will end up being far safer than human-driven cars, in the sense of far fewer accidents. But since the accidents can't be blamed on someone, people will irrationally react to limit/ban self-driving cars. In the end, more accidents, more deaths, but since we can blame a human being - and it's what we're "used to" - it'll win out.
Priorities
In the last 24 hours alone, there have been posts about broken trains on the Orange, Red, and Framingham lines.
Maybe get the MBTA in proper working order before we start devoting resources to what is currently little more than vaporware?
I agree
I was biting my tongue about this but..
We're really wasting our money on this, especially since it's pretty much a notion right now and nothing concrete. Yeah google has 'beta tested' some... but it's not even close to being ready for prime time.
I get seeing the future.. but many of these ideas never come to fruition for various reasons. Let's wait until they are on the assembly line before we start diverting money and resources to this.
Right now we have so much failing infrastructure, not only at the T, but road and bridges. It seems silly to waste money on something that doesn't exist yet.
That and to channel many of my transit friends "We are wasting money on CARS?!?!?"
Not ready for prime time?
Uber begs to differ (well - I guess it depends on your definition of prime time - but we are getting MUCH closer):
http://www.theverge.com/2016/9/14/12900982/uber-self-driving-car-pittsbu...
My definition
Is far different.
As someone who works in IT.. never bother with gen 1 with anything. Prime example.. tons of friends yesterday upgraded to iOS 10.0 . Oooodles of issues. Several friends have bricked phones now and are on their way to the Apple Store to get it replaced. Apple quickly released a small update later yesterday afternoon to fix this.
I wouldn't ride in a driverless car for at least 5 years.. if not longer. Not until they've been on the road a while and have been tested thoroughly. I've worked in software development for far too long to how buggy first revisions can be.
On a side note, this is also the same person (me) who just upgraded to Windows 10 last week. Again, it's best to wait for a few revisions for all the kinks to be worked out.
Agreed, but...
"I wouldn't ride in a driverless car for at least 5 years.. if not longer."
I agree entirely. But what do I do as a pedestrian who wants to cross the street on which one of these things is driving?
Do I even have any way of knowing whether the vehicle that's approaching me is being controlled by a person or a computer?
Yeah google has 'beta tested'
So IOW it's more reliable than the MBTA.
Well
when the MBTA becomes a city agency, or when the T starts investing in autonomous vehicles, this post will be relevant.
Plannig?
From what I have seen, we already have a significant number of driver-less cars.
Granted some drivers are seated behind the wheel but between texting, eating and applying makeup, the cars are pretty much on their own.
zing!
zing!
This has all been foreseen
Easy loophole
Boston drivers are mindless, so this doesn't apply.
I don't know, man
Technology isn't perfect. My computer crashes fairly often. I've had to replace my toaster oven more times than I wish, and that doesn't hurtle down streets at 40 mph.
And when a driverless car goes on the fritz or gets hacked, who gets blamed? At least when there's a human driver responsible for damage or injury, the law can punish the driver [note: that sentence was written in a parallel universe where the law occasionally punishes bad drivers].
There's got to be a better way to get people around town, but I don't think I want this right now.
Yeah, but neither are humans.
Yeah, but neither are humans. The sudden acceleration problems that caused Toyota to recall millions of vehicles were determined to be human error. Toasters may die, but so do people behind the wheel, they fall asleep, get distracted, make poor decisions, etc.
If we wait until the technology is "perfect" we will probably never get there. If we decide it has to be as good a people, we may be there now.
Don't worry
There is a large body of product liability law we can use to determine blame and get people deserved compensation. Not to mention auto insurance. There may be reasons to oppose driver-less cars, that's not one of them.
Yeah, product liability laws
Yeah, product liability laws and auto insurance are two ideas which are perfectly fair, cost-efficient, and provide all the right incentives to increase safety.
Hacked
Cars of today can already be hacked due to their drive by wire tech.
Hacked?
Are you referring to the episode of Dateline where the guy in a lab coat tore apart the dashboard of a Camry, re-routed all the car's wiring through his laptop, and then declared how "easy" it was to hack a drive-by-wire car?
I don't think so
I'd seen an episode or YouTube of a couple of guys who'd done it via laptop. I'll see if I can find it.
OK - there's one from Wired where they killed a jeep on the highway remotely, with the author/journo in it.
You can control failure states
I wrangle bits for a living, so I can talk your ear off about this, but: the short answer to your question is, you can decide what the machine you're designing will do in a failure state. Like, if one of the main cameras goes dark, or the steering wheel stops responding--those are all cases that will be explicitly dealt with, and there will certainly be some sort of failsafe hard-wired into the engine that says "if the fancy driving computer stops responding, cut power to the engine and send an SOS to every other car in range." Your computer crashes the way it does (i.e. a blue screen of death, or a frozen screen) because someone designed it to. Cars will not behave like that.
Meanwhile, the advent of self-driving cars will be the greatest public health boon since vaccinations or penicillin. Seriously. Motor vehicle deaths kill tens of thousands of people a year, and self-driving cars will cut that number by 95%. We owe it to ourselves, or possibly our kids, to do whatever is in our power to get them here as soon as possible.
Well stated
Eric, you are of course 100% correct.
As I mentioned above, the legislation surrounding autonomous cars will not, IMHO, be decided based on science and statistics (i.e. cars will be safer and lives saved) but by emotion. Once a handful of people are killed by self-driving cars (even if it was 100% not the cars fault) there will be a revolt and overly strict laws will be passed.
Riiiight...
/* This should never happen */
Heh, I posted a comment
Heh, I posted a comment before reading this that this will be the exact attitude that we'll see a lot of. Regardless of whether or not self-driving cars end up safer, because can't blame someone for their accidents, people will react irrationally to them.
There already here
Just come to Adams and gallivan and you'll see. Plenty of them
Here's how it'll happen here
Here's another take on the discussion;
Once self-driving cars become sort-of, maybe, just about feasible, (maybe in 5 years?), Boston rejects them for all the reasons listed above (bad roads, bad drivers, bad pedestrians, bad cyclists, and Larry Bird ain't walking through this door any time soon).
But what'll happen, is that Pittsburgh, or SF, or Singapore will have Self-Driving cars. Boston will resist, because, Boston.
Then, the disability community will sue. Because blind people or elderly folks want to be able to get to work and do things. Then a Federal judge will order Boston to form a plan to accept self-driving cars. Then in another four years (see medical marijuana), we'll have policies drafted.
Policies which will state the
Policies which will state the roads will be ready in 10 years at a cost of $5B, yet 10 years out the costs will balloon to $20B and another 10 years... we've heard this story before...
Mayhem
You know they are using GTA V to train the AI for said cars?
http://jalopnik.com/self-driving-cars-will-use-gta-v-to-learn-how-to-dri...
And that means you will have crashes that rival the show CHIPS for ridiculousness:
https://www.youtube.com/watch?v=CqG0hCg5RL4
How is our technology even
How is our technology even close to there yet, when GPSes can't distinguish between the Big Dig tunnel and the surface roads above it, and Boston's traffic lights and lane markings (or lack thereof) are the brain-dead disaster that they are?
I hadn't heard that
I hadn't heard that driverless cars can't reliably detect cyclists. That's disturbing.
Autonomous vehicles only
Make the downtown area autonomous vehicles only! We keep We are continuously reducing the amount of parking spaces in the city anyway...
Marty out front again...to his detriment, again?
Regrettably for the Mayor, he does not have the authority to allow autonomous vehicles to operate on public ways, and until the state makes some rules to allow it, this will be traveling as fast as the F1 cars through the SB waterfront.
The Globe article makes the point only gently.
As long as the computer pays
As long as the computer pays a driving school to help get a road test appointment, it should have no trouble getting a license.
"Boston to start planning for
"Boston to start planning for driverless cars"
What, we're finally going to get proper lane markings?
Great word of financial advice.
I would like to extend to all bostonians some great words of financial advice. Everyone should invest in stock on the driverless cars. Car insurance as we know it will soon become a thing of the past just think of all that money.