Beta
×

Welcome to the Slashdot Beta site -- learn more here. Use the link in the footer or click here to return to the Classic version of Slashdot.

Thank you!

Before you choose to head back to the Classic look of the site, we'd appreciate it if you share your thoughts on the Beta; your feedback is what drives our ongoing development.

Beta is different and we value you taking the time to try it out. Please take a look at the changes we've made in Beta and  learn more about it. Thanks for reading, and for making the site better!

Helicopter Crashes While Filming Autonomous Audi

timothy posted more than 4 years ago | from the send-robot-ambulances dept.

Transportation 218

telomerewhythere writes "A helicopter commissioned by Audi to film its autonomous Audi TT climbing Pikes Peak crashed early this morning. Four people on board were hurt, the pilot seriously. It's a surreal story — a manned vehicle crashes while the one climbing a mountain driven only by computers and sensors carries on. Here's more on the autonomous Audi, a project undertaken with the help of Stanford University."

Sorry! There are no comments related to the filter you selected.

So the weak point in the system is...... (3, Funny)

scosco62 (864264) | more than 4 years ago | (#33626736)

Frighteningly obvious........

Re:So the weak point in the system is...... (1)

nikomo (1338131) | more than 4 years ago | (#33626760)

The fuel lines?

Re:So the weak point in the system is...... (1)

DarkKnightRadick (268025) | more than 4 years ago | (#33626880)

I thought it was unpredictable winds around a mountain.

Re:So the weak point in the system is...... (2, Insightful)

magarity (164372) | more than 4 years ago | (#33627668)

I thought it was unpredictable winds around a mountain
 
While the problem is related to air, it is not the movement of air but rather the lack of it. Pike's Peak tops out over 14,000 feet and the thin air makes a helicopter's responses much slower than at lower levels. If the pilot is having to track the moving car on the ground then his attention was divided, making the situation even more dangerous, so it's easy to understand how a crash could occur.

Re:So the weak point in the system is...... (1)

nametaken (610866) | more than 4 years ago | (#33626954)

Yeah, helicopters. Turns out they're complicated things to keep in the air. :)

http://hardware.slashdot.org/article.pl?sid=10/09/11/2150229 [slashdot.org]

Re:So the weak point in the system is...... (-1, Offtopic)

flyneye (84093) | more than 4 years ago | (#33627004)

Just goes to show, you're only as autonomous as the mystery helicopters behind you watching your every move.
But then they've been following me for years. I think it's Christian Scientologists trying to get me to join Jews for Jesus...

Re:So the weak point in the system is...... (1)

drinkypoo (153816) | more than 4 years ago | (#33627476)

Kind of makes me wonder if anyone is working on manned quadrocopters. Seems like they would be simpler to keep up, at a cost in footprint.

Re:So the weak point in the system is...... (1, Interesting)

Anonymous Coward | more than 4 years ago | (#33627554)

Actually, they should do away with the people, just use a quadrocopter with a proper HD camera, like the Cineflex V14HD [cineflexv14hd.com] for filming and a few more, plus maybe a lidar or two for the remote "pilot." That would significantly reduce the weight, size and cost of the whole system.

Re:So the weak point in the system is...... (0)

Anonymous Coward | more than 4 years ago | (#33627632)

Yea, more moving parts will make it safer... LOL

Re:So the weak point in the system is...... (3, Informative)

ceiling9 (1241316) | more than 4 years ago | (#33627700)

I think the reason is that on a small scale, it is possible to control a quadcopter or quadrotor by changing the relative speed of the rotors, which is much simpler because there is no cyclic pitch, or swashplate mechanism, but this technique doesn't scale well. On a man-sized quadrotor, it would be difficult to accelerate and decelerate the rotors fast enough to have agile control, and so the use of cyclic pitch becomes the better method to control thrust. If you are using cyclic pitch, then it becomes simpler to have one (or 2) rotors instead of 4. Also, when a rotor tilts, it generates large gyroscopic forces. On a small model, these are small compared to the strength of the rotors, but on a man-sized vehicle, the rotors need to be able to "flap" in order to reduce stress on the blades, which again is simpler if you only have to deal with one rotor.

Old news is old.... (-1, Troll)

kidgenius (704962) | more than 4 years ago | (#33626738)

This is from what, Thursday? Friday? Geeze....catch up guys.

Surreal? (4, Insightful)

Karganeth (1017580) | more than 4 years ago | (#33626764)

Since when is a helicopter crash surreal?

Re:Surreal? (-1, Offtopic)

Anonymous Coward | more than 4 years ago | (#33626784)

You just ask the giraffe at the cyclic.

Re:Surreal? (0)

Anonymous Coward | more than 4 years ago | (#33626814)

Since when is a helicopter crash surreal?

When it happens in the matrix [youtube.com]

Re:Surreal? (0)

Anonymous Coward | more than 4 years ago | (#33626820)

Since when is a helicopter crash surreal?

It's more ironic than surreal, but the part that makes it either is spelled out in the words immediately following the dash (where you apparently stopped reading).

Re:Surreal? (4, Funny)

CrashandDie (1114135) | more than 4 years ago | (#33627274)

Except it isn't. Having the riot police interview an elephant based on what the spoon told them with regards to the crash of a helicopter that decided to kill itself during a full moon afternoon because its turbo-girlfriend was dry humping a humvee; now that would be surreal.

This is just a story about a helicopter crash with a few coincidences. Absolutely nothing surreal.

Re:Surreal? (2, Funny)

gEvil (beta) (945888) | more than 4 years ago | (#33626840)

Since when is a helicopter crash surreal?

The helicopter crash itself isn't surreal -- the story is ("It's a surreal story...").

Re:Surreal? (1)

Provocateur (133110) | more than 4 years ago | (#33627546)

Hey now, it'surreal the pilot was seriously injured.

Re:Surreal? (5, Funny)

sakdoctor (1087155) | more than 4 years ago | (#33626854)

Ceci n'est pas un helicoptere

Re:Surreal? (2, Informative)

Anonymous Coward | more than 4 years ago | (#33627248)

Correction: Un hélicoptère

Re:Surreal? (-1, Troll)

Anonymous Coward | more than 4 years ago | (#33627172)

This is not insightful, you morons!

Karganeth is stupid and you're carrying him on a chair. Bunch of...

Meanwhile, good AC comments get zero.

Hey, it would be nice if there was a page with the same content but correct scores...

Re:Surreal? (-1, Troll)

Anonymous Coward | more than 4 years ago | (#33627586)

How ironic, I complain about wrong categorization and get one wrong myself.

Maybe /. is unmanned, too...

Re:Surreal? (1)

Schemat1c (464768) | more than 4 years ago | (#33627538)

Since when is a helicopter crash surreal?

You told me it was surreal!
...It was a pun.
A PUN?!?
No, no...not a pun...What's that thing that spells the same backwards as forwards?
A palindrome...?
Yeah, that's it!

Too soon (2, Funny)

CarpetShark (865376) | more than 4 years ago | (#33626774)

Damnit, I told Audi not to fit Kitt's microlock device [wikipedia.org] before the car was tested against the Three Laws [wikipedia.org] .

Re:Too soon (1)

numbski (515011) | more than 4 years ago | (#33626802)

So this thing will not only drive as stupidly as every other luxury car I see on the road (cutting people off wantanly in traffic, apparently busted turn signals, staying in lanes that end until the very last moment to cut in front of traffic, etc), but now it will also have Super Pursuit Mode? Who's genius idea was this, anyway?

Re:Too soon (1)

neumayr (819083) | more than 4 years ago | (#33626958)

staying in lanes that end until the very last moment to cut in front of traffic?

That may seem like a pretty antisocial thing to do, but if everyone were doing it it would make traffic much more orderly. Where I live, it's actually required by law...

Re:Too soon (1)

mpeskett (1221084) | more than 4 years ago | (#33627254)

Where I live, it's actually required by law...

Where is that? It's a sensible policy... the most efficient way of using two lanes is to use both lanes equally, and merge alternately at the point where one lane ends, rather than mostly all queueing up in one lane and having people merge sporadically all up and down the length of it.

But of course if some people are already queuing, it does seem a tad unfair to blast past them in the other lane, then have to force your way in at the end of it. It only really works if everyone's using both lanes.

Re:Too soon (0)

Anonymous Coward | more than 4 years ago | (#33627382)

Where do you live? Montana?
 
If you stop and think for one second, you will realize the obvious. When the continuing lanes are at capacity, and some jackoff cuts in at the very end, the people in the continuing lanes have to put on their brakes which slows everything down. You get a cascading effect which slows ALL traffic down. It's really nice when some driver wants to be all "friendly" and slows down everyone behind him and lets everyone in the jackoff lane. This is what creates congestion in urban areas. I suggest that you go visit one. And by the way, one should obviously NOT merge at the last moment lest he create a safety hazard where he may have to run off into the shoulder where there may be a broken-down vehicle or debris.
 
When the roads are nowhere near capacity, things are different. This is coming from a guy who consistently drives 25+ over in Atlanta when it's not a rush hour.

Re:Too soon (0)

Anonymous Coward | more than 4 years ago | (#33627662)

You don't know how accurate your are. There was even a study that showed that it was even highly probable that where there was no traffic going home, and you happned to be the guy 'playing nice' that started the chain reaction, by the time you arrive home and watch TV, the traffic report will show the location where you were earlier, now all jammed up. When you mentioned cascading effect, it rang a bell, and they showed the brakelights lighting up; they used the same words to describe it.

Re:Too soon (1)

neumayr (819083) | more than 4 years ago | (#33627958)

One driver being nice has that effect only when people don't use all lanes.
When everyone changes the lane as soon as they find out their lane seizes to exist later on, and only a few people go along the other lane, there is no ordered traffic distribution.
The way it's supposed to work (but doesn't, as it isn't obviously a good idea and people drive by their gut feeling rather by some formal education), is that everyone uses all lanes and take turns at the shoulder: http://data.motor-talk.de/data/galleries/611436/219413/202823618-w250.gif [motor-talk.de]

Re:Too soon (1)

hedwards (940851) | more than 4 years ago | (#33627390)

The problem tends to be the lack of agreement amongst drivers and the tendency of drivers to not let in drivers that they perceive as cheating.

Re:Too soon (0)

Anonymous Coward | more than 4 years ago | (#33627306)

That depends on the amount of traffic. As long as the part with the reduced number of lanes is not a bottleneck, it is more efficient to merge early, because then you can do it at speed. Merging at the last moment usually requires slowing down significantly and that creates a choke point: reduced number of lanes and slow moving vehicles. Conclusion: Merge late in a traffic jam, merge early if there is no traffic jam.

Re:Too soon (1)

ralfmuschall (1782380) | more than 4 years ago | (#33627638)

As others already wrote, there is an efficient way to use all lanes and merge directly before the obstacle. But this is not the question - the antisocial personality of those people enables them to afford such luxury cars in the first place. Here in .de, seeing cars of certain brands (those with "builtin right-of-way") automatically means "CAUTION!". Googling for "Wiehltalbruecke" helps.

Re:Too soon (1)

neumayr (819083) | more than 4 years ago | (#33627860)

Okay, a tragic accident caused by some douchbag on drugs in someone else's expensive car.
Not really indicative of luxury car owner's general behavior, though it's easy to think that way. After all, those expensive cars are made to be conspicuous, it's one of the main reasons the sell ("look at me, I can afford this car!1!"). So, when someone driving this car acts like an asshole, it's more likely people notice and remember it than when someone in his 20 year old Volvo acts up on the Autobahn.
According to my experience, there is no correlation between the car someone drives and their behavior in traffic.

Re:Too soon (0, Flamebait)

Low Ranked Craig (1327799) | more than 4 years ago | (#33626976)

Can't afford one either, huh?

Before jumping to conclusions.... (5, Insightful)

Chicken04GTO (957041) | more than 4 years ago | (#33626776)

Before jumping to conclusions....yeah I know this is the internet...
Flying @ 14,000' elevation aint easy for a helicopter, and it gets *windy* up there at the top of Pikes Peak. Until the NTSB completes the investigation, any comments about what happened and whose fault it is would be pointless.

Re:Before jumping to conclusions.... (5, Funny)

CarpetShark (865376) | more than 4 years ago | (#33626800)

Flying @ 14,000' elevation aint easy for a helicopter

Irrelevant. The helicopter was given a pep-talk before take-off. Neither the confidence nor this determination of the helicopter were factors in the crash, and its endurance was second to none, as it has a lusty wife.

Re:Before jumping to conclusions.... (1)

DarkKnightRadick (268025) | more than 4 years ago | (#33626882)

what?!

Re:Before jumping to conclusions.... (1)

AnonymousClown (1788472) | more than 4 years ago | (#33626904)

He hasn't sobered up from last night yet - or he's drunk off his ass in Hawaii or Australia or he's having a pitcher of Bloody Marys for breakfast.

Re:Before jumping to conclusions.... (2, Funny)

alvinrod (889928) | more than 4 years ago | (#33626966)

Well the editors did say it was a surreal story.

Re:Before jumping to conclusions.... (1)

morgan_greywolf (835522) | more than 4 years ago | (#33626932)

Given, but even supposing that adverse conditions were the cause of the crash, shouldn't the helicopter pilot have known better than to fly at that altitude?

Re:Before jumping to conclusions.... (0, Flamebait)

drinkypoo (153816) | more than 4 years ago | (#33627268)

Flying @ 14,000' elevation aint easy for a helicopter, and it gets *windy* up there at the top of Pikes Peak. Until the NTSB completes the investigation, any comments about what happened and whose fault it is would be pointless.

If anything other than equipment failure or interference with the pilot happened, it was the pilot's fault.

Unless there is any reason to believe that equipment failed or that one of the passengers choked the pilot or something, we know whose fault it was.

Re:Before jumping to conclusions.... (2, Interesting)

Rich0 (548339) | more than 4 years ago | (#33627318)

I'm not big on the whole "pilot error" thing. Too often it just amounts to blaming somebody for the inevitable.

Take 200 of the best truck drivers on the planet. Keep them awake for 48 hours straight. Then set them behind the wheels of big trucks at 2AM and tell them to cover a distance of 800 miles at an average speed of 50MPH or greater.

I can pretty-much guarantee that there will be an accident. Will it be the result of human error - well, sure. However, humans are just another kind of machine. If you took 200 trucks and drove them for 5 years without inspecting their brakes you'd have accidents too - for basically the same kind of reason.

In this case a helicopter at 14k feet is likely at the edge of its performance envelope. That means that any mistake can get out of hand very quickly. Aircraft aren't safe because pilots don't make mistakes - they're safe because they are engineered and operated in a manner that allows recovery if there is a mistake (while preventing the kinds of mistakes that would be unrecoverable). If you operate an aircraft in a way that is at the boundaries of its operational limits, then sooner or later there will be a crash. Sure, most of the time there won't be one, so it is easy to blame the pilot when the inevitable happens.

Re:Before jumping to conclusions.... (4, Insightful)

drinkypoo (153816) | more than 4 years ago | (#33627336)

I'm not big on the whole "pilot error" thing. Too often it just amounts to blaming somebody for the inevitable.

You have a choice on whether you will put yourself in that situation. That makes it not inevitable by definition.

Re:Before jumping to conclusions.... (2, Insightful)

ultranova (717540) | more than 4 years ago | (#33627496)

You have a choice on whether you will put yourself in that situation. That makes it not inevitable by definition.

Oh, sure: you can choose to take the risk and do it, or refuse to do it, get fired and be unable to find a new job because you got a reputation for not doing risky jobs. Given the permanent high unemployment caused by modern technology making more and more jobs redundant, you've better be prepared to live in abject poverty the rest of your life.

Damned if you do, damned if you don't.

Re:Before jumping to conclusions.... (0)

Anonymous Coward | more than 4 years ago | (#33627388)

I genuinely LOL'd reading that. You wrote "Flying @ 14,000' elevation aint easy for a helicopter" and "it gets *windy* up there" in the same paragraph as you wrote "Until the NTSB completes the investigation, any comments about what happened" ... "would be pointless".

Someone call Alanis Morisette (0)

Anonymous Coward | more than 4 years ago | (#33626788)

Someone call Alanis Morisette, sounds pretty ironic to me...

Re:Someone call Alanis Morisette (1)

DarkKnightRadick (268025) | more than 4 years ago | (#33626894)

that stopped being funny about 2 months after the song came out. :p

Re:Someone call Alanis Morisette (1)

pandrijeczko (588093) | more than 4 years ago | (#33627038)

I have just awakened from a 20-year long coma and found it extremely amusing, you insensitive clod!

GPS? (2, Interesting)

Esospopenon (1838392) | more than 4 years ago | (#33626794)

As mentioned in TFA, they hope to create "autonomous driving systems that will one day be integrated into all vehicles as a safety measure". That being the case, I think they still have a long way to go since they have fitted a $100.000 GPS system for guidance. They also have a driver running the course first so the system can "incorporate human reactions", which probably means "learn when to breake" The real test for this system will be when it can cope with unpredictable situations, like traffic lights and old men with hats.

Re:GPS? (4, Insightful)

MakinBacon (1476701) | more than 4 years ago | (#33626816)

There's also the problem that any time somebody using their system gets into an accident, they'll probably try to sue the manufacturer.

Re:GPS? (5, Insightful)

timholman (71886) | more than 4 years ago | (#33626944)

There's also the problem that any time somebody using their system gets into an accident, they'll probably try to sue the manufacturer.

As opposed to suing the manufacturer(s) and/or driver(s) like everyone already does for most car accidents?

The old saw about "We'll never have autonomous cars, because the manufacturers will be sued out of existence after the first crash" is pure nonsense. We already have an enormous amount of computer control in cars, and people are already suing the manufacturers, e.g. Toyota, claiming that those systems malfunctioned after a crash. Toyota is still in business, and the costs of those suits are just folded into the manufacturing costs, as always.

In the U.S. alone, human drivers account for 40,000 fatalities, millions of injuries, and $250B in costs due to auto accidents every year. It would take a pretty unreliable computer system to even get within an order of magnitude of what we do to each other through inattentive or drunken driving. Maybe Microsoft could manage it, but it would be a reach even for them. :-)

When the first autonomous cars hit the road around 2020, what everyone is going to see is the exact opposite - accident rates and costs will plummet. When that happens, auto insurance rates will be adjusted accordingly for autonomous vehicles, and soon you'll find that manual driving is not only expensive, but even illegal in many areas.

Human beings have no business driving. I know this statement bothers a lot of people, but the statistics bear it out. I, for one, will gladly hand over my keys the day I can buy an autonomous vehicle, and never think twice about it. Driving is a chore 99% of the time, and one that I'd be just as happy to turn over to a computerized device as any other chore.

Re:GPS? (1)

MakinBacon (1476701) | more than 4 years ago | (#33627156)

It would take a pretty unreliable computer system to even get within an order of magnitude of what we do to each other through inattentive or drunken driving.

Since when has logic ever stopped people from suing large corporations for things that aren't their fault? If anything, the drop in car crashes will cause them to seem less "normal", and people will be even more trigger-happy with the lawyers.

Re:GPS? (1)

sortius_nod (1080919) | more than 4 years ago | (#33627296)

what sickens me is that your comment rings true.

Re:GPS? (1)

skyride (1436439) | more than 4 years ago | (#33627284)

Human beings have no business driving. I know this statement bothers a lot of people, but the statistics bear it out.

What statistics? Its not like we have a huge pool of data on the number of crashes for autonomous cars driving on normal roads to benchmark it against. Personally, I feel that humans will for a long time still be better driving a car than a machine. Computers are good at dealing with expected information very quickly (trains, monorails, undergrounds, etc) with a small number of variables. But driving a car just has so many potential things that can change, that I genuinely feel it would be nearly impossible to create a computer system that can take into account everything that can change on the roads to thee same level as even a bad driver, at least in the foreseeable future.

Re:GPS? (3, Funny)

morie (227571) | more than 4 years ago | (#33627704)

You want statistics? There seems to be overwhelming evidence that close to 100% of the cars that cause an accident are driven by humans!

Re:GPS? (1)

mestar (121800) | more than 4 years ago | (#33627882)

Humans are better drivers? In what way? Faster reaction time? Better route optimization? Accident avoidance? Lol.

If you could switch every driver with a computer right now, you would double the throughput of roads in every city in the world. With probably one tenth or less accidents.

Re:GPS? (1)

ralfmuschall (1782380) | more than 4 years ago | (#33627738)

> Human beings have no business driving

I agree (I hate it when I have to drive). With robo-cars, there is another problem: Lots of traffic laws are routinely and massively broken. In .de (and probably everywhere), the law fixes a minimal safety distance for different speeds and types of road - and the real distance that the drivers keep is a third to a half of that. This is not just bad behavior - tripling the distance would cut the capacity of the road to a third (unless you triple the speed as well, which would not increase safety either), and there are just not enough roads for that (one would need three times the area for roads as well).

Now, if a robot drives a car, he has two options: Follow the law, cause a traffic jam behind himself (or even provoke somebody to cut into the seemingly extremely long clearance) and be hated by everybody; or drive like everyone else and be sued out of business if something bad happens.

Duh! (2, Funny)

Alef (605149) | more than 4 years ago | (#33627800)

[...] and people are already suing the manufacturers, e.g. Toyota, claiming that those systems malfunctioned after a crash.

Well, obviously the systems are gonna stop working if you crash the car!

Re:GPS? (0)

Anonymous Coward | more than 4 years ago | (#33627380)

This will be solved by making the cars just not right in a way which encourages cheap aftermarket manipulations. (Think jailbreaking to disable an annoying warning or to install a non-braindead hands-free application for an attached cellphone.) The computer systems will log the manipulations and then, in case of an accident, the manufacturer will disclaim all responsibilities. It will then be up to the consumer to prove that none of the manipulations had an adverse effect on the driving capability of the car's computer systems. Can you imagine working your way through hundreds of embedded systems without source code or documentation?

Re:GPS? (1)

CarpetShark (865376) | more than 4 years ago | (#33626852)

I'm curious about the $100,000 GPS system. They sell GPS add-ons for $70. So what kind of GPS costs $100,000? Military, I suppose.

Even if they could make the GPS more cheaply, wouldn't this imply that they expect the very accurate military-grade GPS service to be available to consumers in future? Galileo was abandoned, wasn't it?

Or is this "GPS" actually much more than a GPS -- something more like an aircraft tracking computer?

Re:GPS? (2, Informative)

morgan_greywolf (835522) | more than 4 years ago | (#33626964)

Or is this "GPS" actually much more than a GPS -- something more like an aircraft tracking computer?

Probably they were using something like that. Something along the lines of an inertial measurement unit (IMU) or similar navigational computer. These things usually combine GPS tracking with a precision gyroscope. They can pretty much fly a plane all by themselves, and the military uses them in land-based vehicles, such as autonomous or semi-autonomous tanks. That they could be used to drive a car is not surprising.

Re:GPS? (5, Informative)

mindriot (96208) | more than 4 years ago | (#33627202)

Your $70 GPS addon is way too inaccurate for the kind of autonomous navigation they're trying to achieve. I mean, your standard SiRFstar III claims 2.5 meters of accuracy 50% of the time (a sigma of 3.7 m). That means you can't even be sure whether you're actually on the road, never mind what lane you're in. And that's only in a clear-sky situation. Once you're in a downtown "Urban Canyon" where you hardly pick up any GPS satellites anymore or get wrong readings due to multipath propagation, good luck. Your standard GPS SatNav simply always assumes you're on the road. That won't do for an autonomous vehicle.

You'll need something closer to this high-speed INS+GPS [oxts.com] , the better models of which can be accurate in the decimeter range (assuming careful calibration). The ones I know about are all in the US$50,000 and above price range.

Re:GPS? (1)

usrusr (654450) | more than 4 years ago | (#33627394)

I'd safely assume that the editors were just using the term "GPS" in the way that it is now commonly used by nontechnical people: as a general short for "technomagical gadget that tells cars'n'stuff where to go". They would probably call it a "GPS" even if it wasn't using satellite navigation at all (which it sure does, as the technoligy is just too useful to ignore)

Re:GPS? (1)

Mike1024 (184871) | more than 4 years ago | (#33628002)

I'm curious about the $100,000 GPS system. They sell GPS add-ons for $70. So what kind of GPS costs $100,000? Military, I suppose.

1. A part of the error on GPS is due to things like radio signals slowing down as they travel through the ionosphere. If you set up a GPS base station at a known location, you can take GPS measurements, work out the errors due to the ionosphere (and similar things), transmit that to the receiver on the car, and subtract the errors there. Within a few kilometres of the base station lots of the errors will be common - so a lot of errors are eliminated. (if you don't want to operate your own base station, there are services like Omnistar which operate a network of them and radio out the corrections, for a few thousand dollars a year subscription). High cost receivers support doing this!

2. The GPS signal is comprised of a digital signal with a wavelength of about 300m (which we can measure accurate to about 3m) and a carrier wave with a wavelength of about 19cm (which we can measure accurate to a few mm) - but the carrier signal is a sine wave, so it has an 'integer ambiguity'; 1,000,000.1 wavelengths looks identical to 1,000,001.1 wavelengths. High cost receivers can perform 'integer ambiguity resolution' to figure out the integer number of wavelengths, allowing high precision positioning.

3. There's an encrypted military GPS signal at a different frequency - but using certain tricks you can receive the encrypted military signal. By combining two sine waves using a trigonometric identity, you can get a 80cm sine wave - which means there are fewer ambiguity options, making ambiguity resolution faster. Consumer receivers don't attempt this because you need to receive two GPS frequencies instead of one, and at both frequencies your receivers need ten times the bandwidth.

4. Once you've got high precision GPS, you can put one receiver at the front of your vehicle and one at the rear, giving you a 'GPS compass' which can tell you which way your vehicle is pointing, even if you aren't moving. Of course, using two receivers means paying for two.

5. GPS measurements can be combined with measurements from an intertial measurement unit (IMU) - a sensor system with gyroscopes and accelerometers which can give fast updates, but which are prone to drift over time (as they're based on integrating acceleration to give speed and accelerating speed to give position, a small acceleration error eventually leads to a big position error). The more you spend on your IMU, the lower the drift rate. GPS measurements aren't prone to this drift, but there can be GPS outages (e.g. when going though tunnels) and GPS receivers don't give measurements as fast as an IMU can, so you combine GPS measurements with IMU measurements, usually using an extended Kalman filter.

6. Radio waves can reflect from trees, buildings, and the ground. This is called 'multipath'. High cost receivers use expensive antennas (like choke ring antennas) which have lower gain at lower elevations - which reduces problems with signals reflected from the ground. These antennas are more expensive to manufacture than consumer receivers.

7. Mobile phone GPS chips are produced by the million. The market for high-precision GPS is very much smaller, so the costs of engineering all the above have to be recouped over fewer units - so the equipment is expensive.

In summary, when you spend $100,000 on a GPS system you get a base station and radio link, two rover receivers, all the receivers are capable of receiving the military signal, you get special software that can perform ambiguity resolution for centimetre-precise positioning, you get three high quality antennas, and you get a high-precision IMU and software to go with it.

A lot of this technology could probably be made a lot cheaper if it were mass-produced and installed on every car, as a lot of it's in electronics and software. But in the world of robotics there are a great many sensors that cost $100,000 but could be made a lot cheaper if someone wanted to order a million of them.

If you're interested in autonomous vehicle technology, I advise you to check out The Great Robot Race [youtube.com] - it's a good documentary.

Even if they could make the GPS more cheaply, wouldn't this imply that they expect the very accurate military-grade GPS service to be available to consumers in future? Galileo was abandoned, wasn't it?

Right now, if you're a farmer, you can buy a GPS autopilot system [trimble.com] with high-precision GPS receivers which will link into your tractor's power-assisted steering and guide your tractor down crop rows with centimetre-level precision. Off the shelf, for maybe $40,000.

Galileo's still going - used together with GPS, high precision GPS will become simpler and more reliable.

Re:GPS? (1)

drinkypoo (153816) | more than 4 years ago | (#33627150)

As mentioned in TFA, they hope to create "autonomous driving systems that will one day be integrated into all vehicles as a safety measure". That being the case, I think they still have a long way to go since they have fitted a $100.000 GPS system for guidance.

A "long way to go"? That's the understatement of at least the year. Did you catch this quote from TFA? "If we can design a car that can autonomously go up Pikes Peak, we can design a car that can take over when a driver falls asleep," Kirstin Talvala, one of the students working on Shelley, told the AFP. Wow, no you can't. For one thing, you don't have to deal with other cars when you're making an exhibition run up Pikes. That was a stunningly stupid thing to say, Kirstin.

Re:GPS? (2, Interesting)

mindriot (96208) | more than 4 years ago | (#33627312)

Wow, no you can't. For one thing, you don't have to deal with other cars when you're making an exhibition run up Pikes.

That depends on what you mean by "taking over". If someone falls asleep or has a heart attack while driving, "taking over" can just mean bringing the vehicle to a controlled stop in a safe location and turning on the hazards. In that particular situation, there also won't be much room for suing anyone if something goes wrong -- because had the vehicle not done anything, the situation would've ended gravely anyway.

Also, while you may not have to deal with oncoming traffic running up Pikes, you have to have a damn robust and fast perception system that is able to react to its environment quickly and safely ("oh, pothole on the right, better avoid that"), and you need to have a software capable auf autonomously controlling a vehicle in the most extreme situations (such as going round a curve on a dirt road at 60 miles with a hundred-foot drop on one side. I dare say that if you manage that, you'll be doing pretty well in "normal" traffic as well. Combine this with the expertise gained from the DARPA competitions, and that "long way" is already getting shorter.

Re:GPS? (1)

drinkypoo (153816) | more than 4 years ago | (#33627384)

Also, while you may not have to deal with oncoming traffic running up Pikes, you have to have a damn robust and fast perception system that is able to react to its environment quickly and safely ("oh, pothole on the right, better avoid that")

Potholes don't move. Dealing with moving obstacles is literally orders of magnitude more difficult than dodging potholes.

Re:GPS? (1)

mpeskett (1221084) | more than 4 years ago | (#33627370)

Did you catch this quote from TFA? "If we can design a car that can autonomously go up Pikes Peak, we can design a car that can take over when a driver falls asleep," Kirstin Talvala, one of the students working on Shelley, told the AFP. Wow, no you can't. For one thing, you don't have to deal with other cars when you're making an exhibition run up Pikes. That was a stunningly stupid thing to say, Kirstin.

Well, it's a necessary first step. A student working on a cool project like that, you can forgive a little optimism. Once you've got the thing driving along an empty road on it's own, it's not so very hard to imagine it negotiating traffic too.

Re:GPS? (1)

uop (929685) | more than 4 years ago | (#33627392)

Assuming this is the link to the right article:
http://infolab.stanford.edu/~jonsid/savantcis.pdf [stanford.edu]

Then actually yes the system is aware of parallel vehicles and can avoid such collisions.
Reading some of the article looks like this is more of an exercise in neural networks than anything else.

What scares me most but actually makes sense is that the system does not have any video feed, and as far as I can tell, no feed of objects behind the car. It's strictly forward + side looking, and by "looking" I mean a Laser "RADAR".
An interesting aspect is that all of the "video" information (eg traffic light status) is to be obtained via live feed from "the authorities". This car does not have the sensors to figure out whether a light is red or green. Nor does it have any other video info (citing too costly processing requirements).

Sounds like it could be a pretty good driver assist (sophisticated cruise control + accident prevention), but looks like a truly autonomous vehicle is still some ways off.

Re:GPS? (1)

drinkypoo (153816) | more than 4 years ago | (#33627442)

Then actually yes the system is aware of parallel vehicles and can avoid such collisions.

That's fantastic! But either way the quote is retarded. One problem has significant bearing on the other, but they have substantial non-overlapping problem domains. Perhaps it made more sense in context, but any time you speak to a reporter you have to consider how every sentence will appear out of context...

Surreal (3, Informative)

dimethylxanthine (946092) | more than 4 years ago | (#33626810)

It's a surreal story

No, it's just a clever PR stunt by machines from the future.

Something smells fishy... (2, Informative)

Type44Q (1233630) | more than 4 years ago | (#33626838)

I believe this to have been a joint false-flag operation conducted by Intel, the NSA and VW/Audi, to convince us that we should relinquish control of our Quattros to the machines. I knew there was something suspicious going on when they replaced the five-cylinder with a V6, but no one would listen...

Uber-silly (1, Insightful)

Ancient_Hacker (751168) | more than 4 years ago | (#33626850)

The concept of autonomous autos is just plain silly. There is no way they would ever be approved for use on public roads. Several times a day you make some complex judgement while driving, a judgement that will always be beyond the ability of a computer. Just yesterday: (1) Oh crap, that old lady in the '78 Buick, better give her a wide berth, her eyesight is none too good. (2) A clapped-out minivan full of small kids unloading -- better slow down, they're likely to jump out without looking. (3) In heavy freeway traffic-- what's that ahead, a child crawling across the road?, Nope, looks like one, but it's just a wino's paper bag that slipped off his bottle of wine. No need to slam on the brakes. (4) Whoa, what's that? Oh, of course, nothing to worry about, it's just the shadows of planes landing at the airport a mile ahead.

I don't think your ambulatory computers will ever be clever enough to figure out those situations.

Re:Uber-silly (4, Insightful)

CarpetShark (865376) | more than 4 years ago | (#33626864)

There is no way they would ever be approved for use on public roads.

Oh, they will, but the roads might need a few upgrades. As soon as it can be proven that a car can drive better than a person when the person is trying their best to drive safely, cars will be favoured, since we know people sometimes deliberately drive wrecklessly.

Unfortunately one of the upgrades will probably mean no unpredictable human drivers allowed on the same roads.

Re:Uber-silly (2, Insightful)

thegreatemu (1457577) | more than 4 years ago | (#33626902)

And this is "unfortunate" why, exactly?

Re:Uber-silly (5, Funny)

Marcika (1003625) | more than 4 years ago | (#33626962)

As soon as it can be proven that a car can drive better than a person when the person is trying their best to drive safely, cars will be favoured, since we know people sometimes deliberately drive wrecklessly.

The problem is not with the people who actually try to drive wrecklessly -- it is with the rest of them, the ones who drive recklessly....

Re:Uber-silly (2, Funny)

Anonymous Coward | more than 4 years ago | (#33627012)

...since we know people sometimes deliberately drive wrecklessly.

I think wreckless driving is desirable. Reckless driving.on the other hand tends to result in wrecks.

Re:Uber-silly (1)

selven (1556643) | more than 4 years ago | (#33626884)

I think the 40000+ annual casualties that occur mostly because of drunk drivers, sleepy drivers and distracted drivers (all human faults) are a much bigger problem than the situations you describe. Besides, I don't think the majority of actual drivers ever actually carry out the kind of thought processes you describe in (1) and (2).

Re:Uber-silly (0)

Anonymous Coward | more than 4 years ago | (#33626910)

The concept of autonomous autos is just plain silly. There is no way they would ever be approved for use on public roads. Several times a day you make some complex judgement while driving, a judgement that will always be beyond the ability of a computer. Just yesterday: (1) Oh crap, that old lady in the '78 Buick, better give her a wide berth, her eyesight is none too good. (2) A clapped-out minivan full of small kids unloading -- better slow down, they're likely to jump out without looking. (3) In heavy freeway traffic-- what's that ahead, a child crawling across the road?, Nope, looks like one, but it's just a wino's paper bag that slipped off his bottle of wine. No need to slam on the brakes. (4) Whoa, what's that? Oh, of course, nothing to worry about, it's just the shadows of planes landing at the airport a mile ahead.

I don't think your ambulatory computers will ever be clever enough to figure out those situations.

I think the idea is to have /every/ vehicle AI controlled.

Re:Uber-silly (2, Insightful)

Copley (726927) | more than 4 years ago | (#33626968)

I think you're over-estimating the attention that many people pay to what's going on. For some drivers, your above list would be more like (1) Oh crap, I can't find the track I want on my iPod. (2) Just got to text my BF. (3) How does my hair look in the mirror? (4) Whoops, almost hit that cyclist when I pulled out without looking - better wind down my window and call him a dick.

I think the research into these autonomous systems will never lead to entirely self-driving cars, but instead will lead to driver-assist type systems where dangers ahead (e.g. approaching a corner too fast) that aren't being heeded by the driver, will cause the car to react instead.

Re:Uber-silly (1)

Urza9814 (883915) | more than 4 years ago | (#33626972)

1) If you give all cars proper following distance, that shouldn't matter so much. But either way, just have the computer give more space to any car without the same system installed. I'd assume they can communicate somehow. Failing that, it could probably analyze driving patterns pretty quickly to tell who is driving somewhat erratically.

2) A computer can recognize people, it probably won't have much trouble recognizing small people. If there is a human by the side of the road, go x% slower. If it's a human under y feet tall, go an extra z% slower. Easy.

3) I would expect a computer to be better than humans at not getting thrown off by such illusions. Either way, if it's a serious problem, just add a zoom lens to a camera so it can zoom in on the object in question. Or even better - an infrared camera. A paper bag or scrap of tire isn't going to be giving off heat. A person or animal will.

4) Really? You think a giant shadow is going to throw off a computer? I'm pretty sure that won't be a problem. Otherwise there would be crashes every time the clouds moved. Plus, I would imagine that a digital camera would be able to adjust to changing light levels quicker than your own eyes.

Re:Uber-silly (0)

Anonymous Coward | more than 4 years ago | (#33626974)

The problem is that if there is ever any sort of accident between a vehicle operated by a drunk/distracted/stupid driver and an AI driven vehicle, everybody (judge, jury, cops, insurers) will always give the benefit of the doubt to the human.

Disclaimer IAAPeddler of autonomous vehicles :)

Re:Uber-silly (0)

Anonymous Coward | more than 4 years ago | (#33626978)

so you`re saying a computer might slow down for a paper bag.

is this causing more than the 40000 car deaths per year caused by humans or not?
let alone that > 95% or deaths are on highways. if you only enabled automatic driving on highways where there are no children running around that would account for a lot.

Re:Uber-silly (1)

Punto (100573) | more than 4 years ago | (#33627010)

all the operations you describe sound perfect for a computer to perform. input some data coming from the sensors, output a decision. what's so hard about it? even a human brain can do it

Re:Uber-silly (4, Interesting)

timeOday (582209) | more than 4 years ago | (#33627162)

I don't think your ambulatory computers will ever be clever enough to figure out those situations.

Never say never. It's just a matter of time. Even if some situations are hard to automate, a large percentage of all driving hours (freeway driving, I would think) could be automated much more easily.

The motivation to reclaim driving time is huge. People spend / waste a fantastic amount of time driving. I couldn't find global figures, but apparently Americans spend over 100 hours [about.com] per year commuting (not driving in total - just commuting); the total driving figure in Israel is 577 hours per year [jpost.com] ; and about 40% of mothers in the US spend over 2 hours per day [askpatty.com] driving. Then there are truck drivers and delivery workers whose annual total must be closer to a couple thousand hours per year (i.e. basically their whole life).

Dishwashing machines are very popular, and how much time do they actually save, 20 minutes per day? I can't think of anything the average person more, that could be automated as easily, as driving.

Re:Uber-silly (1)

ichigo 2.0 (900288) | more than 4 years ago | (#33627920)

1 is solved by these autonomous cars (the old lady would not be allowed to drive), 2 sounds like a failure at parking by the minivan driver which would also be solved by an automated car parking in a better spot, 3&4 are perfect examples of why humans should not be driving as their poor eyesight and tendency toward paranoia creates imaginary problems.

Re:Uber-silly (1)

mestar (121800) | more than 4 years ago | (#33627938)

"Several times a day you make some complex judgement while driving, a judgement that will always be beyond the ability of a computer."

You mean like in chess?

"I don't think your ambulatory computers will ever be clever enough to figure out those situations."

And I thing you are already not clever enough to judge that.

Condolences (3, Insightful)

gavron (1300111) | more than 4 years ago | (#33627194)

As a fellow helicopter pilot, I'm happy the pilot and three film-crew members survived.

My condolences to the family and friends of a brand-new (to Air-Cam) Bell 212HP Helicopter.

The world is now smaller by one less helicopter :(

E

Ah That Sucks (1)

Greyfox (87712) | more than 4 years ago | (#33627276)

Hopefully everyone survives their injuries. It's treacherous flying up in the mountains; we have small planes crash regularly up there.

computer-controlled helicopter (0)

FuckingNickName (1362625) | more than 4 years ago | (#33627294)

How many people would have survived the crash had the helicopter been controlled using the sort of tech in the car?

Re:computer-controlled helicopter (1)

mestar (121800) | more than 4 years ago | (#33627926)

There would probably be no crash in that case, so, all would survive.

too bad for the crew (2, Interesting)

quitte (1098453) | more than 4 years ago | (#33627300)

but still I want to see the existing footage now. The teaser clip is pretty cool. Apparently this is not about getting up Pike's Peak but getting up fast. If there are ethical issues showing the helicopter footage at least show the footage from the cars onboard camera that surely exists.
This is so much more exciting than the stupid soccer bots with their Robocup.

The one thing they ARE keeping under wraps (3, Interesting)

Provocateur (133110) | more than 4 years ago | (#33627572)

...is that the car called to report the accident.

Ironically... (5, Interesting)

shrtcircuit (936357) | more than 4 years ago | (#33627626)

We were up on Pikes Peak last weekend staffing a charity hike event when the autonomous car itself also crashed, running off the road somewhere. The wrecker they sent up to fetch it also broke down blocking the road, so they had it shut down for a while getting yet another wrecker up the mountain to help relocate the first one, and get the car out of there.

That thing has some sort of bad omen surrounding it. Everything mechanical around it, including itself, seems to break or crash! I'm amazed nobody has been killed yet, especially with the helo going down on the side of the mountain (that usually ends very badly, so my props go to the pilot for keeping everyone alive).

Physics is a bitch (2, Informative)

Anonymous Coward | more than 4 years ago | (#33627678)

The elevation at the crash site: 13,800ft

Service ceiling of the helicopter 11,150ft

The data is taken from Eurocopter AS355F2, the crashed one was a AS355F1.

Re:Physics is a bitch (0)

Anonymous Coward | more than 4 years ago | (#33627868)

True but the guy flying the helicopter is an extremely experienced high-altitude pilot, you would have to think he had a helicopter outfitted for the task at hand. Maybe not because experience often leads to arrogance but I'll give him the benefit of the doubt until proven otherwise. It's pretty dangerous flying around windy mountains at high altitude even with proper equipment.

Java? (0)

Anonymous Coward | more than 4 years ago | (#33627900)

The video seems to imply the software is written in Java.

What about a driver's license? (1)

tchdab1 (164848) | more than 4 years ago | (#33627934)

Will each individual autonomous car be required to take a driver's test, or will the FCC or DMV or whoever grant USA driving status to autonomous systems as part of their approval process?

And after 21 years, will the car then be eligible to drink and vote (18 years to vote) and borrow money?

Vids (1)

srussia (884021) | more than 4 years ago | (#33628010)

or it didn't happen.
Load More Comments
Slashdot Login

Need an Account?

Forgot your password?