Slashdot Mirror
Blind Man Test Drives Google's Autonomous Car
Velcroman1 writes "'This is some of the best driving I've ever done,' Steve Mahan said the other day. Mahan was behind the wheel of a Toyota Prius tooling the small California town of Morgan Hill in late January, a routine trip to pick up the dry cleaning and drop by the Taco Bell drive-in for a snack. He also happens to be 95 percent blind. Mahan, head of the Santa Clara Valley Blind Center, 'drove' along a specially programmed route thanks to Google's autonomous driving technology. Google announced the self-driving car project in 2010. It relies upon laser range finders, radar sensors, and video cameras to navigate the road ahead, in order to make driving safer, more enjoyable and more efficient — and clearly more accessible. In a Wednesday afternoon post on Google+, the company noted that it has hundreds of thousands of miles of testing under its belt, letting the company feel confident enough in the system to put Mahan behind the wheel."
Disabled man's life put on the line and exploited for corporate publicity stunt.
Boy, if that's not one of the most appropriate metaphors for our time...
Soon, they'll just jack us into our pods, and grow us for the power we generate. :-)
"Flyin' in just a sweet place,
Never been known to fail..."
In the UK you are not allowed to drive unless your eye-sight meets a minimum standard. Is it legal for a 95% blind man to drive in the USA?
For christ's sake, stop talking about the google car! Every time it's mentioned, anywhere, it pushes its release to two years in the future.
Later she was called to the police station to make a statement. The police had arrested the driver. He said he had not seen the crossing because there was thick fog (mildly overcast). Then they discovered that he was registered partially sighted. He had cataracts.
He was convicted of:
His comment to my wife at the police station? "You've spoiled my day". He simply did not realise how serious his offense was.
So I applaud what Google is doing, because I've worked with computers for nearly 35 years, and human beings for over 40, and if the system is designed I would trust the computer over the human being any day of the week, and double on Sundays (drunks with hangovers).
From scarped cliff or quarried stone she cries "A thousand types are gone, I care for nothing, no not one."
"'This is some of the best driving I've ever done,' Steve Mahan said the other day.
I guess he usually uses those pavement reflector thingies to drive by braille.
Driving home tonight there was a young kid playing quite near the road, so I dropped my speed in anticipation of him doing something stupid. He didn't, but I did wonder about the google car making those sorts of calls. I'm sure these google guys are pretty clever and have thought of all these things... are there any video's of self drive cars reacting to these sort of situations?
Like that feeling you get when you see someone else on or near the road and you aren't completely sure that they have seen you and you react by lowering your speed to avoid a potential collision. It's got me out of trouble a few times. If there was an accident you probably wouldn't be at fault, but you've gone one better and seen the accident coming and avoided it.
I'd want to see lots of video evidence of a self drive car doing this sort of thing before I'd be happy sharing the road with one.
Great article. Thank you for publishing this post. Will definetly come back for extra fascinating information.
Cant wait till this project gets shelved in 3 years for no good reason and everyone needs to learn how to drive again...
Did they think of the possibility of driving over a cliff-edge while out of GPS reception?
Or what happens if a bridge collapses? Does the car detect the void underneath it and stop, or just think it's a steep hill and plummet over the edge?
Google cars use only GPS to get direction. The actual driving is done with laser grids and radars (for near distance) and video camera (for long distance). So the car doesn't drive according to what it "thinks" should be there according to the plan, but it drives according to what it "sees" (with its sensors). Any of the situation you mention will end up with the car detecting a lack of drivable surface, stopping, asking its GPS for an alternative route and going another way.
Wherever a human driver will react the same, or will be too busy getting distracted with on-board enterteinment/smartphone/passengers/news paper, etc. and fall into the hole is left as an exercice to the reader. (yup, we've already had stories on /. of clueless drivers wrecking their cars because the GPS told them to go a certain way).
There are a BILLION and one problems, that only happen once in a lifetime.
And that's why you put the stuff on extensive testing. Already in the range of hundreds of miles on actual raods with actual traffic in the case of Google cars.
Yes, it won't take into account some really weird exceptions. But... Humans make mistakes too, and mostly in normal everyday boring situations (because they are boring and the brain kicks into "autopilot" routine mode). (And chance are that some of the really weird situation are going to be "missed" by the human, not because the human wouldn't have had reacted correctly, but because the human was to bored to pay attention). Also even if you, the human driver, think of hundreds of situation which might be missed by an IA, but where you think you'll be able to react correctly, I can probably think of situations where you drove perfectly well, but still got into an accident because of some other driver.
At some point in time, we will reach the situation where an autonomous car (even including the accidents due to weird rare situations) will cause a lot less casualities than a human driver (who might just not be paying attention).
The point of this publicity stunt is to show that, given the current extensive testing the cars have undergone, this point in time is nearing soon.
And, also, the advantage of autonomous cars is, as the wierd situation happens, they can be analysed and the programming can bu updated, leading to even less casualities on the long term. .
Whereas, with human drivers, you can't just magically "programatically remove from the road" asshole, idiot and dristracted drivers
You can't verify a system on this scale. It's like trying to verify a Kinect. You just cannot guarantee what it will detect something as just by a simple test of something similar. And this is orders-of-magnitude more complex, more important and more deadly than a stupid games console.
You can't prove *mathematically* that the autonomous car will be perfect in absolutely every single situation (juste because there is a potential inifity of such situations). But you can prove *statistically* that the autonomous car is better and safer than a human driver based on the number of accidents and casuality caused by both. And overall this *will* mathematically increase the safety on roads.
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
The drive-up ATM's at Citibank branches in the NY area have had Braille labels on all the buttons for years. Seemed kind of silly up until now, you certainly would not want a blind person standing in the driveway using the ATM, and I certainly hope a person requiring Braille labels on an ATM would not be behind the wheel of a car. Not knowing Braille myself, I always assumed the labels said "Get out of the way!!! You're standing in a driveway!!!".
But now I realize that Citibank was preparing for the eventual release of the autonomous car.
"We make our world significant by the courage of our questions and by the depth of our answers." Carl Sagan
As explained by other, the car *does* slow down, and even eventually halt when exposed to situation it thinks it can't handle.
Also, the car has much lower reaction times. So in some situations, it doen't really need to slow down, it will react immediatly if needed, whereas a human driver will need to slow down to make room for slower refelexes.
(The distance between autonomous vs human-driven cars on the motorway, for example).
"Sufficiently advanced satire is indistinguishable from reality." - [Tips: 1DrYakQDKCQ6y52z6QbnkxHXAocMZJE61o ]
NFL Jerseys are popularly received in many countries. People crazy about NFL events tend to buy NFL Jersey for collection or gifts. NFL Jerseys just set off overnight. Because of people’s preference, there being a growing demand for NFL Jerseys. Moreover, it boosts market for NFL Jerseys. We design, sell and deliver various NFL Jersey Wholesale around the world. Our products are of good qualities. The USA authentic nfl jerseys we designed are durable. We have our own website. And the price we provide for Football Jerseys are reasonable enough that you can rely on us without any hesitation. If you are interest, don’t hesitate to contact us through the website where more details of NHL Jerseys are provided. And it is recommended that you order the NFL Jerseys online. We assure you that your order of our Discoun NFL Jersey NBA Jerseys NHL Jerseys MLB Jerseys is of good value for money
NFL Jerseys from http://www.cheapjerseysaleshop.com/cheap-148-NFL-Jerseys.html
...Steve Mahan != Steve Mann.
(Note to Google: a similar test with Steve Mann has the potential to be really, really interesting.)
I did read some where that there was this case where the gps did not make it clear that you needed to stop and wait for a ferry and some one did not stop and the car went in.
what about GPS having you drive on restricted roads? roads that are gated or cut off but the GPS shows that it goes though or having a truck go down a road that is no trucks or dumb stuff like sending a truck down the tail of the dragon.
I don't understand most of these what ifs, you act as if they're unsolvable. My *PHONE* can detect a one-way sign through it's cheap camera and a simple app, these cars will have a wide variety of sensors and more powerful computers. In addition there are standardized signs for no trucks, restricted, do not enter, etc... The car can easily read these signs (OCR could even be employed) and simply route around it. If a car wanted to make the next left, happens across an orange detour sign, or cones in the way, it will instantly recalculate and just route around it. It's really not that hard to imagine these solutions.
We know how stupid humans are, how could we create anything smarter than ourselves?
I would think you would build upon the intelligence of hundreds if not thousands of people, much like all of the technology our civilization is based on, or do you happen to know how to manufacture everything you use from scratch including microchips, displays, etc.?
" do you happen to know how to manufacture everything you use from scratch including microchips, displays, etc.?"
Nope - but I don't have them running anything important without human intervention either.
I don't know how to build a car either, but I'm the one driving it.
FWIW -- I agree that a computer driven car is going to be far safer than anything operated by a human (or monkey, snail, slug, etc). I was just trying to point out why people may inherently be against it.
... for the baby boomer generation to age out of driving. It will now be much easier for all those seniors with upside-down mortgages on houses in developments where you must drive to buy anything - they can stay in their homes.
I can certainly see this working well on highways and other types of roads that are long, uninterrupted, and highly predictable. Especially if there's special lanes for them, like the HOV lanes but smarter. (Insert driver intelligence quips here.)
I don't know how well they would do on the residential streets with all the variables in place. Add on the need for any type of sensors or other fixed equipment and having them off the highway is in a far more distant future.
Would I want them driving in regular traffic? Not so much simply because of the mix of automation and humans. The autonomous cars simply can't cope with the number of variable reactions that humans provide. Not to mention the soon to be developed hobby of messing with the autonomous cars to see how they'll react. I can't believe it wouldn't happen. Then the blame would be on the car, not the idjit who decided to see what would happen if he slammed on his brakes in front of one when traffic is going 50mph.
There's a future in this. Eventually. And every advance is a good one, imo.
The interesting part is if we can get past the knee jerk reactions when the autonomous car has hit its first car, has been hit by other car due to confusing behaviour, has killed its first pedestrian, has gone over its first cliff killing the passengers etc because accidents will happen. Accidents will have to be investigated as thoroughly as air plane crashes.
The technology seems pretty wonderful, but like all things made by men, it will at some point fail. Even if the software is designed to handle all possible scenarios and implemented without flaw (good luck), it's still all dependent on physical sensors which are subject to physical failure, sometimes in subtle ways.
Inevitably, if our society permits these to be widely used, we'll see one crash into something (or someone) somewhere. What then?
Who is at fault for the crash? The person behind the wheel? The company that built the hardware? If we write law to the effect of "the person in the driver's seat is always the driver and responsible for the car", that ixnays the blind-driver scenario pretty fast -- too bad for Mr. Mahan! If we choose to designate the manufacturer of the autopilot as the responsible party, expect claims against them Real Fast -- "No officer, it wasn't me that ran over that kid, Google was driving!"
I understand it's not black-and-white -- I favor responsibility distributed to the "driver", but I have personally had my brakes fail dramatically due to improper maintenance by the auto dealer. I would have disclaimed responsibility if I'd crashed after my brakes failed.
However the it shakes out in the courts, it seems sensible to deal with the legalities before the systems become popular -- if they are tested enough, it becomes possible that they'll be widely deployed before the first Real Incident occurs, which will only serve to make it more dramatic. That's drama I could do without.
Will Google foot the insurance bill if I use their gear?
My only complaint was that I was not allowed to "drive" this thing. I realize that "driving" it is about as involved as being a passenger in any other car, but man would I want to do it anyways.
"The car can easily read these signs (OCR could even be employed) and simply route around it. If a car wanted to make the next left, happens across an orange detour sign, or cones in the way, it will instantly recalculate and just route around it. It's really not that hard to imagine these solutions."
actually OCR is more or less setup with DOT signs (and its a US federal level standard) so combine that with a bit of pattern recognition and we would be set (you would need patrec to account for seeing half of say a stop sign and knowing that it is a stop sign.
Any person using FTFY or editing my postings agrees to a US$50.00 charge
I think that generally, automated vehicles would indeed be safer than your typical human motorist. However, there is one situation that gives me reason to hesitate; carjackings.
A year or so ago, there was a group of thugs in East St. Louis Missouri who, upon the approach of a single vehicle in the 2-3am hours, would run into the road to block the vehicle, and when it slowed, would beat and rob the owner. It's our natural instinct to stop when we see someone in front of us; and even in this case it's arguably the correct response (I suppose it's better to be beaten and robbed than to have the guilt of hitting someone on your conscious). Any autonomous system would be programmed to do the same; making it trivial to predictably stop the car just by approaching it from the front. When this situation happened to my friend, rather than slow down, he downshifted and sped towards them. They jumped out of his way. It was a risky move, but he avoided being robbed and possibly killed.
My (perhaps unwarranted) fear is that when such systems become common, these kinds of crimes will become routine. Even moreso if/when the driver is unable to override the collision detection. And yes, I realize that this is an outlier case, but you know how people tend to give unnecessary weight to outliers that affected them personally.
A lot of the resistance to the idea of automated vehicles is 'who is responsible if there is an accident'. If one ever does get in an incident, especially the first couple of times, I expect the payouts to be higher than for an equivalent human-caused incident. I think it is easy to measure the safety performance of driverless cars, and I expect that it is better than regular drivers, and will improve. That means that insurance for driverless cars should be cheaper than for human drivers. So Google should offer to insure all their driverless vehicles, and because those vehicles are safer, they will come out ahead.
I want to know how did he know what was on the Taco Bell Drive-in. Don't tell me the car read him what was on the order board!
Paul: Father... father, the sleeper has awakened! - Dune
I want to see the CVs of these engineers first. On how many high-integrity systems have they worked so far? I know plenty of people with an AI background, and trust me, I don't want these to program my car. I'd also need to know which programming languages and development tools they have used, see the source code, and would like to know which formal software and hardware verification methods were used to verify the code.
I'm fortunate enough to know some of the people on the team. They come from robotics backgrounds and are very experienced in highly reliable systems. For example, a big chunk of the Google team was hired out of the Carnegie Mellon Tartan Racing team. That team tested their systems so much that they wore out hardware. The Carnegie Mellon team also originates from the Field Robotics Center which sends robots to Antarctica, volcanoes, sinkhole lakes, the mountain deserts of Chile, and abandoned mines. These are hardened systems developed using hard core systems engineering and serious software engineering methods. Trust me, this is not a group of ivory tower AI people ignoring version control and using latest beta release of LISP. One of the technical leads, Chris Urmson gave a talk last year at Carnegie Mellon. It was apparent that they have continued these practices in Google. The incredible volume of field testing is the outward evidence.
On a related note, I always thought traffic congestion must waste an enormous amount of fuel. Years ago I came across a stat related to this done by a Texas university, which stated that the daily average was something like 250k barrels of oil per day - which may seem like a lot, but bear in mind the US plows through ca. 9 million barrels of gasoline per day.
My original link is now a 404 but I've found the data source they used: Congestion Data for Your City — Urban Mobility Information. Go to the data for "All 439 Urban Areas," and down to the table for sums. This states that due to traffic congestion 1,943,330,000 gallons were wasted for 2010, thus 5,324,205 per day; there are 42 gallons in a barrel, thus 126,767 barrels/day. I'm not sure if this is the correct way to calculate this, as refineries actually crank out more like 19 barrels of gasoline per full barrel of crude; so maybe 280,221 b/d? That must be how the original study came to its conclusion. Either way it's not much of a dent in the 8.71 million barrels of oil we wolfed down in the form of gasoline last week: Petroleum and Other Liquids - Data - U.S. Energy Information Administration (EIA)
It will never happen in first-world countries, for one simple reason: litigation. Even if automobile fatalities were reduced by 50%, all it takes is one dumb kid running out into the street, causing an accident that neither human nor robot could avoid. Then the manufacturer and every other company in the supply chain will be sued for millions upon millions. Lawyers will have a field day. Insurance rates will skyrocket. Companies will refuse to shoulder the liabilities, so they won't even make them--it's just too financially risky. Even if they are statistically safer, that won't matter in court when an accident does happen (and they will, because nothing and no one is perfect). A human driver is responsible for his own driving, but the manufacturer will be responsible for its robotic driving, and no car manufacturer with lawyers in their "right" minds would allow such a thing to proceed.
Maybe it could work in other parts of the world, places with less-developed legal systems....
"Those who consume the bulk of goods are those who make them. We must never forget this secret of our prosperity."
well soon in GOP USA people in lockup get healthcare care with fully covered vs working 39.5 hours a week with no plan on the out side.
autopilot software goes under FAA review and they just don't let any piece of code go into the code base.
even long term there will be a switch over period with auto cars and non auto cars on the road at the same time and likely there will be manual drive only areas and or areas where you can't wall off pedestrians. (maybe you can but very few areas have las vegas pedestrian overpasses)
I'm surprised there's no discussion of some of the most interesting points. Namely, if automatic cars are commercial and commonplace, then who's really in charge of them? Considering the potential danger of any errors in any changes to the control software, I'd have to expect it would be protected from any changes every way you could imagine (TMP, signed bootloaders/OS, all of that). Or else some inept hacker/terrorist/total nutcase will screw it up and cause some huge wreck.
So Google or Ford or whoever will probably be the ones in control of the software, and therefore the ones really in control of what the car does. And large companies tend to do whatever the Government asks them to, because they have a lot to lose. So, how long until it's set up so that if the police want you for any reason, good or bad, any car you get in will take you directly to them, refusing to do anything else? Probably all of the cars will also keep records of everywhere they go, and maybe everyone who rides in them too. And all those records will be in the hands of Google or whoever, available to any police agency with, or maybe without, a court order. Maybe available in bulk, for whatever bulk analysis anybody can dream up. Want a list of everyone who drove by that street where they sell drugs? Everyone who drove by a crime scene, just to see if there are any witnesses or accomplices? Here you go. What else can you dream up that someone might do with all of that data?
I do see the tremendous number of lives that could be saved and injuries prevented when human error is taken out of driving. I also think it's worth considering what else we might be giving up by doing it.
I don't reply to ACs
Tons of lame excuses in this thread for not accepting the start of a great technology. I have been saying for years that I wish my car could hook up to railroad tracks and autopilot me along them. This system would be so much better. Most of the complaints are from not RTFA. I know, must be new here. The rest of the complaints most humans cannot even deal with. The amount of energy that it would take to be as alert as these sensors 100% of the time would probably kill a normal human or at the very least, the humans senses would quickly degrade. I think a great start to this would be patches of straight, uncluttered, highway that is designed to handle this type of vehicle. Go from there! Sign me up!
In Mountain View, California, Google car drives you!
I believe Google presently relies on hyperaccurate maps (previously collected laser scans) of the route. It seems unlikely that they would go through the trouble of collecting this data if the route was inappropriate. In the future, it seems likely that they'll want the cars to adapt to roads that have never been traveled before, but how often is it that you happen upon a road that lacks Street View coverage (which seems like a great "vehicle" for collecting the laser scans)? Even if self-driving cars only work on (a) highways, (b) major arteries, and possibly (c) routes that you've manually driven before, that's still a huge net win, even if it means I have to take control whenever we leave the self-driving routes.
How does the car know to stop at the drive-thru speaker? How does it then know when to go again and pull up to the drive-thru pay window? And same again for the pick-up window? And later in the video, how does it know that it's time to choose a parking space? So clearly there is some human control, none of which could be done by someone who is actually fully blind? So in this case, was the 95%-blind person doing something (gas/brake, voice commands), or was a sighted person controlling it in some way? Basically, how does a driver (sighted or not) communicate decisions to the car that are beyond the scope of "driving"?