Self-Driving Cars Will Be In 30 US Cities By the End of Next Year

schwit1 sends this report from the New York Observer: Automated vehicle pilot projects will roll out in the U.K. and in six to 10 U.S. cities this year, with the first unveiling projected to be in Tampa Bay, Florida as soon as late spring. The following year, trial programs will launch in 12 to 20 more U.S. locations, which means driverless cars will be on roads in up to 30 U.S. cities by the end of 2016. The trials will be run by Comet LLC, a consulting firm focused on automated vehicle commercialization. ... they’re focusing on semi-controlled areas and that the driverless vehicles will serve a number of different purposes—both public and private. The vehicles themselves—which are all developed by Veeo Systems—will even vary from two-seaters to full-size buses that can transport 70 people. At some locations, the vehicles will drive on their own paths, occasionally crossing vehicle and pedestrian traffic, while at others, the vehicles will be completely integrated with existing cars.

Comments

Boston, in the winter?

[QuietLagoon]

I'd like to see one of those self-driving cars find its way around Boston this winter....

Re:Boston, in the winter?

[Peter Simpson]

I'd like to see one of those self-driving cars find its way around Boston this winter....

Top post. Nice. Boston says, "bring it on!". Commuting into Boston is not for the faint hearted. I've seen potholes this winter into which the Google self-driving car would fit nicely.

Re:Boston, in the winter?

[jeffmflanagan]

If the roads are really as bad as you say, maybe Google should deploy self-driving hovercraft in Boston. A self-driving car should be able to spot and avoid potholes at least as well as a human driver.

Re:Boston, in the winter?

[QuietLagoon]

Will the Google cars be smart enough to evade the college kids crossing the street against the red light while buried in their smart phone displays?

Are google cars smart enough to stop at an intersection that has a green light, but also has a policeman with his hand up telling the car to stop?

.
Did you know Google's self-driving cars can't handle 99% of roads in the US?

Re:Boston, in the winter?

[C0R1D4N]

This could be solved by a handheld the device for policemen to hold up that the cars can easily recognize, the same device could be used by construction companies (or built into new slow/stop spinny signs)

Re:Boston, in the winter?

[JeffOwl]

The cars should be smart enough to stop for any object blocking the road or moving predicted to be blocking the road when the car gets there. The question of the police officer is interesting though. And what if it isn't a police officer? What if there is simply a car or truck stopped blocking your lane, loading or unloading or just double parked. Will the car know when it is safe to cross into the other lane and go around?

Do the cars have an alarm or something that alerts the occupant (back-up driver) in the event it gets confused?

Re:Boston, in the winter?

[ITRambo]

Where are the self-driving flying cars when you need them?

Re:Boston, in the winter?

[jader3rd]

I'd like to see one of those self-driving cars find its way around Boston this winter....

How well are the humans doing at it?

Re:Boston, in the winter?

[QuietLagoon]

The cars should be smart enough to stop for any object blocking the road...

A policeman in the center of an intersection is not blocking either lane of the road.

.
Also, around here, there's a habit (started long, long ago, before there were cars) of putting flagpoles in the center of main intersections.

Re:Boston, in the winter?

[AK Marc]

Sometimes it was hidden by the vehicle ahead of you, sometimes it's in a part of the road where you'd have to swerve into an oncoming lane to avoid it,

If you can't see a road hazard because you are too close to the car in front, then back of. Yes, I'm aware of the "but then, someone may pull in front of you" argument. If you are driving so close to the person in front that nobody can pull in, you should stop driving.

Are these potholes so wide they can't be straddled? I did a quick search online and the pics of Boston potholes aren't as bad as claimed here. Sure, there's an occassional outlayer, which gets lots of attention, but they don't look that special. Oh, and why aren't you going to http://www.cityofboston.gov/pu... and reporting them every time you see them?

I think they go under-reported because Bostonians want something to complain about. So you get what you ask for.

At least with the self-driver-car, the car can report the pothole when it starts, possibly avoid small one, to avoid making them into the big ones. The Boston drivers here talk like they aim to hit them as hard as possible, for the badge of another bend/broken rim.

Re:Boston, in the winter?

[AK Marc]

The article was hilarious. "Google cars can't drive on roads.*" "*update, Google says everything we said is wrong." "**further update, MIT says Google is sugar coating it."

There's no reporting, no investigation. Just an opinion piece presented as an article, with updates that contradict it at the end, without editing the original at all. All that supports is "people fear the unknown."

Re:Boston, in the winter?

[AK Marc]

You are directly contradicting the example given, then saying the example is bad. If you want to do that, please re-state it as a new example. If not, your comments are all a non-sequitur.

Re:Boston, in the winter?

[oldmac31310]

Most definitely coming in the year 2,000. Please wait patiently.

Re:Boston, in the winter?

[AK Marc]

No. Seeing a green light, then ignoring it for a human signal is a much smaller edge-case than a human of any kind directing traffic.

Sort of like the difference between a traffic light that's off and one that's covered in a burlap bag. One means "4-way stop" and the other means "green light". And most people don't know the difference, or even that there is one. So why are you holding the computers to a higher standard than the people?

For people, 95% of the time, they do what the person in front of them did. Even in violation of law.

Re:Boston, in the winter?

[smooth wombat]

Are these potholes so wide they can't be straddled?

It depends. Sometimes you can, but sometimes the way they have grown they're just large enough for one wheel or the other to drop into. If the pothole is on your right, you could try to swerve right to straddle it but you might also hit the curb or parked cars.

If you try to swerve to the left, you might miss it, or end up perfectly hitting the target, but as the OP said, you run the risk of driving into oncoming traffic.

As to reporting potholes, every major city responds to reports of potholes needing patched, but the sheer amount of requests, traffic, time of year, etc, prevent them from quickly filling them. When you include limited monetary resources, things get much worse.

If you have never driven in colder climates where potholes are ubiquitous this time of the year, you don't have a frame of reference to understand how bad these things are. It's like people who have never lived down south don't understand the perpetual heat and humidity combined with insects left over from the dinosaur era roaming about.

Re:Boston, in the winter?

[The Grim Reefer]

Will the Google cars be smart enough to evade the college kids crossing the street against the red light while buried in their smart phone displays?

If not, they can document it as a feature. They can call it "Evolution in Action".

Re:Boston, in the winter?

[AK Marc]

As to reporting potholes, every major city responds to reports of potholes needing patched, but the sheer amount of requests, traffic, time of year, etc, prevent them from quickly filling them. When you include limited monetary resources, things get much worse.

Boston claims 2 days. Anchorage claims 1 day. I remember the last time there was an "incident" with potholes in Anchorage (worse weather than Boston, but Boston complains more). It turned out that people had documented the pothole's growth and size for weeks, but nobody reported it. It's a form of "my hardship is worse than your hardship" one-upmanship.

If you have never driven in colder climates where potholes are ubiquitous this time of the year, you don't have a frame of reference to understand how bad these things are.

The AK in AK Marc stands for Alaska. Care to talk about my experience in colder climes some more?

Re:Boston, in the winter?

[WhatHump]

How about Sudbury, Ontario, Canada? This past winter a "pothole" (more like a sinkhole) ate a resident's Ford F150. We set a record here in Ontario for coldest February in recorded history. Roads are buckling everywhere here in Toronto, and there have been so many watermain breaks some people have been without running water since January.

Re:Boston, in the winter?

[LinuxIsGarbage]

If you lived in MA this winter, you'd know that you don't always get the option to avoid potholes. Sometimes they eat your car. Cue the cracked rims.

You can always spot the ones who don't live in cold climates by that they assume that you can always spot the pothole in time to avoid it. Sometimes it was hidden by the vehicle ahead of you, sometimes it's in a part of the road where you'd have to swerve into an oncoming lane to avoid it, and sometime it is filled with just enough snow to look flat but not dense enough snow to keep your car's tire from falling into it, etc.

I live in Eastern Canada. We've had a particularly rough winter this year, but even normally, there's a lot of freeze thaw cycles that causes havoc on the roads (creating potholes). I'm used to driving anticipating potholes. Changing lanes because I *know* there's a pothole ahead (from the last time I went down the road). Or picking which side I straddle a pothole on based on potholes I know are ahead. Or controlling by braking on an offramp or at an intersection based on which sections are rough, and which are smooth. Can Google do this? Add some radar, and some reports from previous vehicles and hopefully it can. Especially adding when roads suddenly go from dry pavement to whiteout low traction conditions.

Even with the best attempts at straddling the potholes, at night, add some mild (near freezing) conditions where road brine is spraying all over the place, and it's hard enough to see where the road is going, let alone if there's any potholes you don't already know about.

Re:Boston, in the winter?

[LinuxIsGarbage]

As to reporting potholes, every major city responds to reports of potholes needing patched, but the sheer amount of requests, traffic, time of year, etc, prevent them from quickly filling them. When you include limited monetary resources, things get much worse.

If you have never driven in colder climates where potholes are ubiquitous this time of the year, you don't have a frame of reference to understand how bad these things are. It's like people who have never lived down south don't understand the perpetual heat and humidity combined with insects left over from the dinosaur era roaming about.

We get to laugh at them down south when they get 1" of snow and hundreds of vehicles are written off in wrecks, and a state of emergency is declared.

With potholes cities are hesitant to patch them while it's still freezing out (cold patches don't bond as well to the road), but will patch really bad potholes on major roads.

A lot of potholes go under-reported. I'll drive by a pothole for a couple weeks, send in an online work request, and within two business days the hole is patched. This is even on major roads where you'd think someone from the city (like the 4 busses an hour) would drive by. Yet rather than send a work request to the city, people rather complain to the local radio station's stupid morning show's "find the worst pothole in the city". Same with dead streetlights. Send in a request and it's fixed.

I've called in and reported pathways that weren't fully plowed days after a storm, and another plow is dispatched.

I was impressed when I called in to report an open manhole cover on a major road (which could cause serious damage if not straddled), that the operator responded "Is it at such and such address? That one has been reported". People other than me do call in problems on the road.

Re:Boston, in the winter?

[AK Marc]

In every state I have ever lived in, a traffic light that is clearly 'off' and one covered in a burlap bag both mean '4-way stop'.

In Texas, whether by law or convention, a signal covered in a burlap bag is no longer a signal. It's obviously not a broken signal, as it's not like the wind blew in some snow that covered the lights, but a worker, working on the ight, deliberately obscured it.

In practice, when this is a 4-way stop, there will be signs along the road to control the traffic. But for new crosswalks, it's common for the light installers to put up the lights weeks or months before the traffic engineers have programmed them into the city-wide light controlling system so they can be activated. So they'll sit, covered in bags, for months, as a green light. Then, one day, the bags will be gone, and the lights will be lit. And if they go dark after that, they are a stop sign.

Re:Boston, in the winter?

[JeffOwl]

I guess I didn't do a good job of separating the two issues. 1st issue from the GP: avoiding kids crossing the street - Car should be able to do it. 2. Police officer, regardless of where he or she is standing at a given instant, gets to direct traffic. I would hope the car wouldn't run over them no matter where they are standing (see issue 1) but the real question was more to whether the car, when self driving, would be able to understand what the officer was telling traffic to do. And further, how to deal with other less common traffic situations, like double parked cars, broken down cars, construction areas, etc.. The car may be smart enough to avoid running into things, but will it be able to negotiate its way around the anomaly?

Re:Boston, in the winter?

[pspahn]

Check out this guy! His infrared vision allows him to see road stripes through several inches of slush and ice. You should be a fucking movie star you godsend from beyond.

Re:Boston, in the winter?

[Peter Simpson]

"Limited" monetary resources? Hah! Try "nonexistent"! Boston's budget for snow removal disappeared faster than the snow came down.

And two days is about right for the time it takes to fill a pothole...and the time it takes one to reappear after being filled! Talks about a thankless job. The roads are so heavilty travelled, they have to do it at night, when it's coldest.

There isn't even a central location to report the potholes, which are repaired by the individual towns, or the state, depending on who is responsible for that section of the roadway.

I once head about an app that sensed sudden Z-axis accelerations and used the GPS location to report potholes, but I have never seen it and it was probably a dream someone had.

Cheer up! Spring is on the way! I can see the grass over my septic tank again. The foot and a half of snow everywhere else and the six foot banks are going to take till June, though...

Tampa is not Tampa Bay

[Anonymous Coward]

Tampa Bay sounds like a likely place for a self-driving car to end up, although a self-driving submarine might be more useful.

Sounds good

[gurps_npc]

Please note that this test includes buses - which are far more likely to become the first self-driving vehicle that a private car.

The vehicles travel slower, set routes. The cost to add the self-driving capability is a lower percentage of the total cost of the vehicle. Finally, over the long term they save money by removing the necessity of paying a driver.

Still not as perfect as using the tech on garbage trucks. They move even slower, have less union opposition (because you are only getting rid of the driver, not the attendants that load the vehicle. But no one's perfect.

Re:Sounds good

[Anonymous Coward]

They have already automated trash trucks to be rid of all but the driver and sometimes one attendant. Here is an example: http://www.gosanangelo.com/new....

Re:Sounds good

[Anonymous Coward]

I seriously doubt that. One word: Unions.

See Toronto's Scarborough RT as a reference. This train, built decades ago on a dedicated track, was fully capable of running fully automated yet they never managed to remove the driver primarily because of pushback from the transit union.

Re:Sounds good

[N1AK]

All the reasons you give for why buses apply to subway systems, trams, trains, and planes. Often they are even stronger reasons to automate those tasks but yet most aren't yet... A bus has to be able to handle interactions with lots of passengers which adds considerable complexity. I'd expect trucks will be some of the first automated vehicles. They get loaded up at one depot, drive to another depot without needing to stop or enter dense urban areas, then stop at another depot. Bus drivers also tend to be paid notably less than lorry drivers (no idea why given the similar/identical license requirements). If a lorry full of goods is at risk of crashing it can make selfless decisions because their is no risk to life, which is a considerably easier legal situation than a bus that has to decide between killing a pedestrian vs risking the lives of dozens of passengers.

Re:Sounds good

[CaptSlaq]

"The vehicles travel slower, set routes."

IOW Asian cars.

Well, yaknow... at least they can negotiate a corner.

Re:Sounds good

[Em Adespoton]

I seriously doubt that. One word: Unions.

See Toronto's Scarborough RT as a reference. This train, built decades ago on a dedicated track, was fully capable of running fully automated yet they never managed to remove the driver primarily because of pushback from the transit union.

And to counter that, look at Calgary and Vancouver. The trick is to do it with new infrastructure; replacing the old infrastructure will take a LOT longer.

Re:Sounds good

[ShanghaiBill]

They have already automated trash trucks to be rid of all but the driver and sometimes one attendant..

Where I live (San Jose, CA) there is only a driver. The driver uses a joystick to control a robot arm which grabs and dumps each container. If you position your trash can where the robot arm can't grab it (like putting it too close to another trash can), then your trash doesn't get picked up that week. This feedback ensures that people rapidly learn to put out their trash cans correctly. We have some of the highest labor costs in the world, so if something is automated anywhere, it is automated here.

Re:Sounds good

[NMBob]

What if someone gets on the bus and doesn't pay? If a rider has a heart attack is the bus going to cal 911? If you try to force one of these buses/cars off the road with your vehicle will it fight back?

Re:Sounds good

[OzPeter]

They have already automated trash trucks to be rid of all but the driver and sometimes one attendant. Here is an example: http://www.gosanangelo.com/new....

Um, you do know that this sort of technology has been around for at least 20 years?

Re:Sounds good

[ShanghaiBill]

Bus drivers also tend to be paid notably less than lorry drivers (no idea why given the similar/identical license requirements).

A major reason (at least in America) is that most lorry drivers own their vehicle, and pay their own taxes, insurance, benefits, etc. So you aren't just paying for the driver, you are also paying for the truck, and all the expenses that go with it.

Re:Sounds good

[CrimsonAvenger]

The driver uses a joystick to control a robot arm which grabs and dumps each container. If you position your trash can where the robot arm can't grab it (like putting it too close to another trash can), then your trash doesn't get picked up that week.

I recently had occasion to be awake insanely early and was shocked to discover that we have the same sort of thing here. I would never have suspected it, have never heard anything about it on the news.

It's possible that the vehicle I saw was an isolated case, though. Perhaps they bought one or two to try them out before making a larger purchase (or not).

Re:Sounds good

[meta-monkey]

We've had them in my semi-rural town for a few years now. My 2 year old is infatuated with trucks, and garbage day is like christmas. He hears the truck down the street and rushes to the window to watch it rumble by and pick up cans.

Re:Sounds good

[MrL0G1C]

If you try to force one of these buses/cars off the road with your vehicle will it fight back?

This is of course a most common problem these days, it's a rare occasion when I get on a bus without someone trying to force it off of the road. Not to worry though, these buses can transform into giant killer robots.

Re:Sounds good

[AK Marc]

They are very common outside the US. I can't recall ever taking a manned airport tram outside the US. In the US, they don't want the automation. If a driver makes an error, they blame the driver, and hold the company blameless, so you can't sue. Computer-driven trams are safer, but would get more lawsuits in the US. Computer-driven isn't used in the US for liability, not technical issues.

Re:Sounds good

[NMBob]

Oh good! That's what I was hoping they would turn into. It'd be pretty boring if they just flashed you the finger on the little sign that has the route number. :) I was really thinking more about all of these auto-autos. With our lack of adherence to the three laws, so far, it could get interesting out there.

Re:Sounds good

[RespekMyAthorati]

We had them in SLC 25 years ago.

Re:Sounds good

[RespekMyAthorati]

A major reason (at least in America) is that most lorry drivers..

We don't have lorry drivers in America.
We have truck drivers.

Already got 'em

[OffTheLip]

Based on the number of drivers I see texting while driving there appears to be an abundance of self-driving cars where I live.

Is Veeo Systems real?

[chadkennedyonline]

Why does a Google search for "Veeo Systems" not return a link to the company? Only returns references to the company. Something seems fishy.

A real test: Orlando, FL

[msobkow]

If they want a real test, try Orlando, Florida. I found it the most trying city to drive in of any I've ever lived in, thanks to the joyous combination of people visiting from Ohio that expect a mile clear ahead of them and people from New York who think 6 inches is enough of a gap for someone to cut them off.

Re:A real test: Orlando, FL

[CaptSlaq]

If they want a real test, try Orlando, Florida. I found it the most trying city to drive in of any I've ever lived in, thanks to the joyous combination of people visiting from Ohio that expect a mile clear ahead of them and people from New York who think 6 inches is enough of a gap for someone to cut them off.

Baby steps.

Re:A real test: Orlando, FL

[GLMDesigns]

If they want a real test, try Orlando, Florida. I found it the most trying city to drive in of any I've ever lived in, thanks to the joyous combination of people visiting from Ohio that expect a mile clear ahead of them and people from New York who think 6 inches is enough of a gap for someone to cut them off.

Common, six inches isn't enough room to fit the nose of the car in. Everyone knows that you need at least 14 inches to inch your way in. :-)



Truthfully - once you get past a half a car length you're asking to be cut off.

Re:A real test: Orlando, FL

[taustin]

You've obviously never driven in Atlanta, where every other street is Peachtree something-or-other. Peachtree Street, Peachtree Road, Peachtree Lane, Peachtree Circle, Peachtree Blvd, Peachtree Way, Peachtree Up Your Butt. The locals think it's funny.

(Also the only city I've ever been to where I saw a uniformed motorcycle cop pick up a hooker on his department bike, but that's another story.)

Re: A real test: Orlando, FL

[don.g]

As a tourist from New Zealand, the toll booths are my overriding memory of Orlando. Well, that and the Kennedy Space Centre.

The really fun thing to do is drive from the wrong airport to the right one, running late for a direct flight to SF (and then across the Pacific, home - not a flight you want to miss). By the third or fourth toll booth I was flinging correct change at the attendants.

Re:A real test: Orlando, FL

[pspahn]

The suburbs around Denver have a lot of the same names of streets that exist in the city core road grid.

North/South streets will usually be ordered alphabetically and with a theme. Plant names, city names, historical names, etc. This is great, it makes a lot of sense because you know if you are on Ivy then you know the next block over is Holly.

Now, in the suburbs, they decided to do away with a grid road system and instead went with semi-random twists and curves and such. The grid is gone, but the names of the streets remain the same. On top of this, you have several streets with the same name, differing only on the suffix as you mentioned. Ivy Street, Ivy Lane, Ivy Road, Ivy Avenue, Ivy Court, etc. So while the guys in the city decided that putting Ivy next to Holly next to Jasmine etc. the folks that laid out the streets in the suburbs also thought that was a great idea, so good that they made everything intersect with everything unpredictably.

Fuck you Aurora.

They will be there, but not in great numbers.

[eepok]

We are still a very, very long way off from any attempt at saturating American roads with driverless vehicles. It won't happen until we have decided (at the legislative level) how liability is to be handled and set some very specific guidelines for human take-over of the driverless vehicles.

I'm still trying to figure out which will hit the mass market first: battery-swapping EVs, hydrogen fuel cell vehicles, or driverless vehicles. I have a feeling it's in that order.

so quiet

[swschrad]

all you hear is the occasional "thump-thump." occasionally muted screaming if somebody gets caught in a fender.

How will be the first to be killed?

[BrendaEM]

Volunteers, anyone. Do you feel safer?

Re:How will be the first to be killed?

[SecurityGuy]

No, I don't.

Then again, dropping my kid off at school this morning I saw one parent almost crash into another at 5mph because they weren't paying attention. It's not like I feel safe now.

Re:How will be the first to be killed?

[jader3rd]

Volunteers, anyone. Do you feel safer?

Given how dangerous humans are, yes.

Now for self-governing countries...

[mi]

Self-Driving Cars Will Be In 30 US Cities By the End of Next Year

Now, if only we could have a self-governing country...

(Please, don't hate.)

Why cities?

[jfengel]

Cities seem to me like the worst place for automated driving. They're not great for any driving, since things are constantly coming at you from all directions. And while computers are great at operating with many simultaneous distractions, these are cases where errors get people hurt or dead. Erring on the side of caution will block traffic, and city streets are often already at capacity.

I would think that the best use for automated cars would be interstates, which have limited access and more predictable situations. Problems turn into crises fast, but that's the kind of thing where a computer could react better than a human, since it's likely to involve less fine discrimination between "human" and "non-human".

Ultimately I'd love to see automation replace all human drivers in cities, since it can break the connection between driver, destination, and necessity to park. They could coordinate more effectively at intersections, which are currently very wasteful. So I'd like to see this work, but right now it feels like begging for trouble.

Re:Do not want.

[Coren22]

As opposed to buses/cars driven by humans? Whenever there is an accident, everyone lawyers up. The nice thing about the automated system is that there are recordings of everything; they will clearly show that the bus was cut off and had no choice but to rear end the car that cut it off.

Re:Driverless cars - questions

[93,000]

Very good questions. Likely answers below:

Yes, but you'll probably get some sort of ticket for doing it if you get caught.
Yes, but you'll probably get some sort of ticket for doing it if you get caught.
Yes, but you'll probably get some sort of ticket for doing it if you get caught.
Yes, but you'll probably get some sort of ticket for doing it if you get caught.

Eventually as the cars get proven I would guess that regulations will loosen up, however right away I'm sure we'll be expected to keep fully alert while 'not-driving'. I predict that it will be 'all the responsibility, twice the boredom' of normal driving initially.

Differences

[DrYak]

How would an accident by a self-driving car be any different than one controlled by a human but caused by a mechanical malfunction?

Human with failing car:

• Adaptability: Living being are known to be highly adaptable and creative. There's a chance that, encountering an unexpected event, the human driver will react and manage to save the day by pulling some weird manoeuvre and avoid causing any casualty. That's different from an AI where an engineer must be sure to have taken into account every possible situation, and also be sure that the car will fail safely/take correct actions even in the course of an unplanned event. Bascially you need to engineer the adaptability that comes for free with a real flesh driver
• Liability: at least there's a clear first line of responsibility with a car: the driver - the one responsible for the car. Then it's up to the driver to decide to sue the car manufacturer or the car dealer if it seems that it's more a car defect than a piloting error / failure to properly maintain. Whereas with a truely autonomous car (with no steering wheel) you can bet that there is going to be huge legal mess about whom to blame

With cars you get:

• Cold blood: a computer will always follow it's programming, no matter what. Whereas a human being might panic and freak out instead of act accordingly and might cause more damage/casualties due to innapropriate reaction (e.g: a person less used of slippery roads can panic and be unable to stop the car on icy roads and lose control of the car), a car will stick to the plan. If the engineer has correctly planned ahead and has correctly provided a safe failing strategy, in the event of danger the car will follow the plan and manage to keep the damages low (e.g., although not an autonomous car: in the rare few accident involving a critical battery failure, the on board computer has given a warning well in advance and the driver has been able to pull the car to the side, exit and be at a safe distance well before it burst into flame. Similarily, if an autonomous car senses something really problematic, it might be able to pull out to the side, stop in a safe spot, and eventually call for help).
• Concentration: Not directly incident-related, but preventing them. A human has a limited concentration span. More and more stupid useless things are competing for the small mental capabilities that our brain can provide (people texting on their smartphone is a common cited exemple here on /. There has even a crowd funding project to create a voice- and gesture-controller HUD for cars which among other advertised twitter connectivity). With machine, it's just easy to throw more dedicated computing power for each task that needs attention. A semi-autonomous car with adaptive cruise control (like lots of high-end car nowadays) will always be watching ahead, no matter what. A driver will probably glance quickly in the side mirror before doing a manouvre. A car's blinds spot system is constantly looking back/using sonnar all the time. etc.
  To go back to the mechanical failure: a driver might be too distracted to notice early since of imminent failure and it might be too late to react. A car's computer will always be controlling tire pressure. An collision avoidance system will always notice when the truck in front is losing a big heavy object and will to an emergency breaking on time
• upgradeability: when an incident happens (or even better, when a human managed to successfully avoid an incident), it's possible to analyse it, and create an upgrade that will better handle similar situation in the future. Such upgrades can potentially even be made available instantly over-the-air to current owner of similar cars (Tesla is known to propose such uploads).

Re:Differences

[meta-monkey]

Yeah, but when the robot uprising begins, they've already got you.

"Doors locked. New destination: processing plant. Have a nice day..."

Re:Differences

[AK Marc]

To go back to the mechanical failure: a driver might be too distracted to notice early since of imminent failure and it might be too late to react. A car's computer will always be controlling tire pressure.

My wife has driven her can for 3 months with a near-constant "intermittent error" (I've seen it on about 90% of the time, she claims it's on about 10% of the time, given the amount of time I'm in her car, that's statistically possible, but highly unlikely). The self-driving car can drop her off, then drive to the dealer. The self-driving car will be safer because those little things can't be ignored, so mechanical failures should be lower.

Also, self-drivers will have near-constant communications with "home" (near-constant being either real-time, or batch when stopped, or batch when stopped plus real-time for "incidents"), so they can report things with vibration sensors and such.

I had a friend pick me up to go somewhere. As we were driving, I put my hand on the dashboard, paused, and said "I didn't think you had a full-spare in this car." He was confused. I said "You recently changed the right-front tire. But I didn't see a steel-rim on it, so I presume you have the flat in the trunk." His only response was "bullshit." He thought I talked to his parents or something. He didn't think it possible to tell from the passenger seat that something was wrong, then put a hand on the dash and tell which of 4 tires was recently changed. He later told me I was 100% right on all counts. I was seriously interested only why Chevy had their Impala SS spares on full-alloy rims. I'd have guessed that they'd use a steel rim, even if full-sized. I have no idea if it was an extra-cost option to get the allow-rim spare. It's not like they needed a donut to save space in the Impala SS trunk.

The spares are usually balanced poorly (they aren't used that often), or are properly balanced, and fall out of balance over the years in the trunk. So a vibration from the right-front was detectable by a human, even if most wouldn't notice or know what it was if they noticed.

A few vibration sensors in a car, correlated with mechanical failure reports would probably diagnose a large number of problems, long before they happen. And cut repair cost, as problems could be identified early, when small repairs would save a larger bill later.

As for failure modes, I've heard you are more likely to die by trying to avoid a deer than hitting it without slowing. Doing nothing is better than trying to not hit it. A human would never take that action unless they were too drunk/tired to have a slow response time. The best action is drive straight and brake. But humans don't like that either. Human's responses are slow, and usually wrong. A computer-car would be better in almost ever case, but people will focus on that 0.01%, rather than save 30,000 lives a year by moving to self-driving cars. The other thing is that the more self-drivers are out there, the safer it is for everyone.

Run to the hills!

[MagickalMyst]

I don't know which is scarier... (some) human drivers or self driving cars.

What, with all of the viruses, "hacking" and snooping these days... The thought of some 13 year old kid hijacking cars remotely from his/her parents basement...

Running to the hills sounds like a good idea... oh, wait! There be drones in them hills.... sigh.

'caution slow moving vehicle'

[MrL0G1C]

You could take the shuttle, or if you're in a hurry, you could run!

https://www.youtube.com/watch?...

Re:Do not want.

[Immerman]

No, I suspect liability will rest largely with the maker of the autonomous driving system - you know, the ones responsible for actually driving the bus. If I recall correctly we already have Google and a few of the others volunteering to accept liability.

Re:Do not want.

[Immerman]

I fail to see how the autonomous driving system producers would be any less "in the cross hairs" than a driver would be, if anything they'd be more so, in large part due to the "scariness" of the new technology.

A driver can have a bad day, suffers from normal human fallibility, etc. We all understand that. A computer does exactly what it's been instructed to do, always. If it's responsible for an avoidable accident then it's because the designers failed to consider something important, or simply decided that their algorithms had been refined to a level of acceptable risk (which, if the risks are 1/10thas high as for a human driver, is completely respectable - nothing is risk free, and engineers, like insurance companies are in the business of quantifying acceptable risk.)

The bus service may still be the "target of first resort", but it's going to be damned hard to fault them for using a system with a well-documented history of being considerably safer than a human driver (unless there's a competing, even safer autonomous system). And the first rule of lawyering is "sue the people with money". At present that's the bus service, because regardless of fault, the driver is just a working stiff. Google though - they probably make the bus service look like paupers. And especially early on Google et. al. will likely even be offering some degree of indemnification - something along the lines of "if one of our properly maintained autonomous systems is involved in an accident, then our legal team will help with your defense, and we'll cover N% of any penalties levied against you". (And if it wasn't properly maintained then the bus service *should* be roasted - just like if they made a habit of hiring chronically drunk drivers.)

Time marches on (we're in trouble now)

[iq145]

1) http://www.newser.com/story/18... 2) http://hardware.slashdot.org/s... 3) http://www.newser.com/story/19... 4) http://www.dailydot.com/techno...