Digital Cameras Help Alert Sleepy Drivers

An anonymous reader writes "An interesting story on how digital cameras are being mounted in cars to watch the eye movements of drivers to make sure that they are awake. The cars include two cameras, one watching the road and one watching the driver. If there is something on the road that is a danger and the driver doesn't see, the car alerts the driver. Pretty neat technology."

Comments

Comment removed

[account_deleted]

Comment removed based on user account deletion

Cool intermediate technology

[Dancin_Santa]

Obviously, the final application of this kind of technology is to allow the car to take primary control of the vehicle and let the passengers relax in peace.

We already have navigation systems that are accurate to within half a meter in many cities worldwide. We also have collision detection algorithms (aka hashing functions) that can help avoid crashing into other cars. We now can mount cameras onto vehicles to provide visual sensory input.

All we really need is an IR sensory input for fog driving.

In cities, this kind of "decide the destination" driving without the hassle of actually driving the vehicle would be really useful, I think.

Re:Cool intermediate technology

[Goosey]

In cities, this kind of "decide the destination" driving without the hassle of actually driving the vehicle would be really useful, I think.

Great idea! We could call it Taxi!

Re:Cool intermediate technology

[xtrvd]

As much as I would like to believe that there are people striving to make automatic automotive technology available tomorrow, I have a sneaking suspicion that some people out there who make a living off of a class 4 license will argue that this cannot and will never replace a human.

Re:Cool intermediate technology

[Colin Smith]

"All we really need is an IR sensory input for fog driving."

Just stick on an IR camera and cars will be able to drive themselves? Nope, we're decades away from fully automated vehicles. Real roads are far far more complex than the test roads which they have been run on so far.

http://robotics.eecs.berkeley.edu/~janka/PATH/st er eo_drive.html

If you want fully automated vehicles right now, a segregated guideway is required, AKA Personal Rapid Transit.

http://faculty.washington.edu/jbs/itrans/

Re:Cool intermediate technology

[Dave_M_26]

knowingly breaking a speed limit

major cause[s] of death and destruction on the road

Whilst I agree that some breaking of the speed limit is obviously dangerous (e.g. going >60mph in a 30 zone) I disagree that a strict adherence to the speed limit is necessarily good or safe.

For instance, going 40 (in a 30 limit) on a clear straight road, on a bright Sunday afternoon is probably going to be safer than doing 30 on a rainy Monday morning, down a winding road in front of a school, despite the fact that the speed limit is nominally the same.

I would be wary of any system which was incabable of taking these varying factors into account (not, in theory, difficult - weather sensors, tyre grip sensors, visibilty distance monitors, pedestrian detectors etc).

Dave

Re:Cool intermediate technology

[Colin Smith]

"Two major causes of death and destruction on the road - gone."

Speeding isn't a major cause. 7% at most, in fact 80% of accidents happen within the limit. Far more important factors: Drunk, tired, distracted (eg mobile phone) and plain stupidity.

We could have fully automated vehicles today. Just not on the roads, they are too complex. There are already systems which can do the job:

http://www.skywebexpress.com/
http://www.atsltd .co.uk/

Re:Cool intermediate technology

[Proc6]

I would think just an additional lane appended to each side of major interstates would be plenty. I dont think we really need automated pilots to drive 3 blocks of residental roads to a McDonalds. But a single, controlled-environment, automatic-pilot-only lane on I-80 would rock and feasible far sooner. Something with no bikes, no motorcycles, just cars equipped with special equipment. Maybe a verification booth at each end of long stretches to make sure people entering have said equipment.

Re:Cool intermediate technology

[Colin Smith]

The 3 city blocks to McDonalds are the most dangerous roads, interstates are relatively safe. They've had automatic driving kit for interstates for quite a while.

e.g.
http://www.cvhas.org/

They use magnets embedded in the lane to determine position. The issues, what happens at the exit if the driver falls asleep? How do you handle unexpected situations like wildlife on the motorway? Who's at fault when an accident does happen, the manufacturer?

The other thing is that it's a relatively expensive and inefficient way to apply IT to transport, a kludge even. All the vehicles (millions of them eventually) would have to be retrofitted with kit, all the motorways would have to be retrofitted for it to be effective, it's an expensive and rather slow proposition.

Privacy concerns

[Goosey]

I can see the obvious saftey benefit from this, but perhaps the possible privacy conerns should be considered.

Suppose this follows a logical step and they add a link to a centralized server that monitored traffic volume to help the results be more accurate.

Suppose insurance companies were able to gain access to data this could produce, and started factoring your on-road alertness into their rate

Yano on second thought, that doesn't sound that bad at all.

Re:Privacy concerns

[tedu]

good. :) my rates will go down because i won't have to subsidize all you bumblenuts who can't pay attention.

Re:Privacy concerns

[polecat_redux]

Suppose insurance companies were able to gain access to data this could produce, and started factoring your on-road alertness into their rate

And why shouldn't they? If you drive half-asleep/drunk/retarded, your rates *should* go up, and insurance companies should be able to access any relevant info about your driving habits in order to determine your rates.

It's getting to the point where simply being inconspicuous with deviant/dangerous behavior is no longer sufficient to avoid the consequences, and I say good. I think that if you drive drunk, or speed, or drive erratically, you should get a ticket regardless of whether or not a cop happens to be present at the time. And yes, I'm talking about equipping cars with devices that can detect such crimes. Too many people confuse this with an issue of privacy or civil rights, but I don't believe it to be. Such a thing would merely serve to lift the veil of obscurity that many people tend to hide behind as they threaten the lives of those around them.

Re:Privacy concerns

[ambrosen]

Well, driving safely except for the fact that you were overtaking where there wasn't enough room.

Re:Privacy concerns

[ambrosen]

Good point. So at what speed does a modern car have a braking distance of 14m (3 car lengths), then? Why separate licence for motorways when they're by far the safest roads. How does lights on at all times improve safety outside of a heavily forested country where light is in short supply in winter and not as bright as expected in summer evenings where dark cars are preferred (I assume Sweden was your reference point for this)? Predicting dangers is part of the driving test. UK government stopping distances are based on a Mk 1 Ford Cortina or similar.

And if you're actually worried about losing your licence when a head on collision's on the cards, your priorities are very wrong. Why so important to overtake anyway. Seeing as you're talking about the UK, it's not as if you've far to go on your single carriageway road. And they are by far the most dangerous roads by quite some margin.

Re:heat waves?

[LiENUS]

I would imiagine it would be a false positive. Which is much better than a false negative, false postive makes you a little annoyed for a short while. False negative sends you out of controll into a ditch, your decision.

Digital camera kept me awake!

[eingram]

I used to drive a lot for long hours during the night. I would catch myself dozing off quite a bit..

I'd then break out my digital camera and take pictures of the road, myself, buildings, etc. I'd also set it on my dashboard and do a long exposure image to catch the headlights of cars and city lights (for a cool streaking effect). I had a lot of fun and it kept me awake.

Was it dangerous? Nah. I can operate my camera without looking at it really, so I was able to keep my eyes on the road (and keep them open).

Alerts you to dangerous things on the road?

[Seumas]

If there is something on the road that is a danger and the driver doesn't see, the car alerts the driver.

Now, if only they can devise a way to keep 85 year olds who think that it's their god given right to drive until the day they die, from slamming on the gas and destroying buildings and killing pedestrians because they thought it was the break pedal - or driving into THROUGH AN AIRPORT because they thought you return your car at the Hertz inside the airport.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Wonder how it determines what is a danger

[LiENUS]

I wonder how this determines what is a danger and whats not, does it detect any sudden change in the road such that a pothole or cones on the side would set it off or is it more specific in that only if you go off the road it works?

Better or Worse?

[N Monkey]

I can't remember the exact figures but I heard that in the UK either "1 in 6" fatal accidents may be caused by falling asleep at the wheel. Certainly they've been advertising the dangers of driving while tired as much now as anti-drink-driving.

Now I can see it could save a life if a so called "micro sleep" occured at the wheel but could it have the opposite effect? Would some people then try to drive longer thinking they have a safety net/alarm clock to wake them up if they drift off?

Re:Better or Worse?

[tristan-jt2]

I remember seeing a documentary on TV a few years ago, can't remember if it was in the UK or in France.

They had taken a dozen of drivers fitting a particular set of criteria. They had to be used to driving at night, and drive a certain mileage every week.

They rigged them up with monitoring equipment, set a couple of night vision cameras in the car and sent them on their way.

They had to drive a distance that was estimated to take about 4h.

Once at the destination a team of boffins would ask them to fill in a questionaire about how they felt about the drive, whether they felt tired, etc.

The results were simply scary:
None of them had bothered to take a break.
None of them declared having felt tired enough to feel they had to take a break.
Yet their brain activity was showing numerous periods that looked like deep sleep for less than 5 seconds.
On average these periods amounted to a whooping 6 minutes over the 4h of driving.
On the videos you could just see the drivers blinking for a unusually long time.

Having a system that detects that I'm blinking in a suspicious way, gets the driver seat to vibrate, and then sound an alarm if I don't open my eyes immediately would certainly not annoy me. I'd take the hint that I need to take the next exit and try to grab 1/2h of sleep.

My sister fell asleep at the wheel once while on the motorway and told us that she had only blinked, only to open her eyes after feeling what she described as a bump.
She took the next exit because she was feeling seriously tired and slightly puzzled about the "bump". Turned out she had hit the safety rail after drifting all the way over the fast lane.

Re:Better or Worse?

[343 Guilty Spark]

As far as the opposite effect, this is a know phenoma in driver safety;

It is called risk compensation or behavioural adaptation, if you look at things like accident rates before and after seat belts have been made complusory you find that while fatalities for those in cars decrease it is not at the level predicted by statistics, and fatalities for pedestrians and cyclists acutally increase..

Basically people feel safer so they drive faster/aren't as careful.

The interesting thing is that the effect is strongest for safety changes that make the chance of an accident lower (e.g. ABS breaking) and weakest for things that just reduce the cost of an accident (e.g. Airbags).

I my own research I have found that even in simulator studies that risk compensation appears to happen as far as road width goes (i.e. if roads are widened people drive faster, if they are narrowed they drive slower).

Re:Better or Worse?

[Dave_M_26]

Basically people feel safer so they drive faster/aren't as careful.

I always thought that an interesting experiment would be to remove the driver's seatbelt and fix a large spike to the steering wheel. I suspect the number of accidents would go down ;-)

Dave

Complacency

[tod_miller]

I do not mind having some driver alert system, but if it goes off too many times, too many false positives, then drivers may ignore it, yet at the same time, the very fact that it is there might make them more willing to distract thier attention from the road.

So it gives you a false sense of security, but like all computer equiptment, you ignore it the seconed it gets too annoying.

How many times has a car alarm gone off, and you rush outside to apprehend the thieves?

This sounds too much like a tax funded project gone awry. Perhaps the car might have a failsafe mode if the triggers go too far? if the person doesn't hit an ok button in enough time, the car should slow calmly and require some special intervention to make sure the user is aware.

Now any action on the part of a computer that would remove the human from the loop is not desirable, as this would mean a car might slow in the middle of a 5 lane intersection, or something stupid.

But if humans take themselves out of the loop through complacency, then that is worse.

This is great and all..

[kagaku]

..but shouldn't it be the drivers responsibility to stay awake while driving? If you're tired enough that you need a camera to watch your eye movement to make sure you aren't falling asleep, should you really be driving?

Repeat with me

[Anonymous Coward]

Social problems require social solutions, not technological band-aids. The reason we have so many sleepy drivers in the first place is a combination of bad urban planning that results in millions of people taking extra-long commutes, one person to a car, and erosion of labor rights that makes it possible for employers to overwork their employees and tire them.

It wont work!

[Goth Biker Babe]

There's this psychological effect called risk compensation. It's been shown that the safe people believe they are the less careful they will be. So if you have mechanisms in your car to stop you from being stupid you'll actually be even more stupid that you would normally be and so the whole thing balances out.

Here in the UK Volvo drivers have a bad name with motorcyclists. Why? Because they are very safe cars and so many Volvo drivers take less care than someone in a less safe car. But cars aren't the only thing on the road and it's all well and good you being safe in your car if you're involved in an accident but what about the other poor sod!

Actually the best thing to make everyone drive safely and wear seat belts and the like is to put a spike in the centre of the steering wheel!

Re:It wont work!

[richie2000]

I totally agree. The trick would be to make people safer while making them believe they're in mortal danger. It's a game of visualising risks. Since they can see they're going to be two minutes late for that important meeting but can't see the broken down truck over the next hill or see themselves as pieces of decaying flesh at the roadside, they put the pedal to the metal.

And this is why bikers in general (there are sadly a LOT of exceptions to this rule) are not as prone to accidents as people in general (the non-biking public, as it were) might imagine. We're too close to the road to not notice it rushing by at break-neck speeds. I believe it's best said in Zodiac where a bicyclist, all dressed up in black, is asked why he doesn't have any lights or flourescent clothing and he responds with "For that to work, I'd have to assume every motorist around me is wide awake, sober and not trying to kill me. That's stupid. I pretend there's a million dollar bounty on my head and everyone's trying to hit me. It's my responsibility to make sure they don't." and there's a certain amount of truth in it. That and the spike works for me. :-)

Citroën already does that (sort of)

[MartijnL]

The new Citroën C5 (http://www.citroen.co.uk/) has a lane departure warning system that detects if the car is leaving it's lane (like if when the driver has nodded off and there is a bend in the road). It only warns when the driver crosses the white line however so collision detection is still a way to go.

Bad move

[Underholdning]

Pretty neat technology
What? This is not neat. This is a step backwards. If drivers gets accustomed that their car will alert them if they're about to hit something, it will probably encourage them to contiue that extra hour of unsafe driving.

Re:French cars *are* innovative

[Arminator]

French cars *are* among the most innovative cars.
Especcially Citroen seems to be at it. They used Headlights that shine into curves depending on how far you turn the steering wheel already in the 60s or 70s.
They were the first to use a very soft suspension (advertised with a car bumping over a freshly plowed field with raw eggs on the backseat. The eggs remained intact).
They built a hygropneumatic suspension that automatically stiffens the suspension. For example you drive a car speedily over a long hump, and the inertia lifts the car upwards, while the road begins to go down again. Now imagine a curve right after that. With a soft suspension the car will swing around with a lot of load-cycle changes, while the hygropneumatic suspension stiffens and keeps the car steady.
This in turn is an evolution of their suspension that adjusts the height of the car's rear so that you can easily load the trunk in a lowered car, and when startinging, it lifts the back up again compensating the load in the trunk.
Furthermore a lot of automobile companies (Opel, Volkswagen among them) used Peugeots Diesel-motor technology, since it is among the best engineered Dieselmotors.
And a few years back Peugeot made the HDI Diesel engine, that produced very high exhaust heat, so the carbon particles get burned, eliminating the black smoke from Diesel engines.

So, when do you think did the French cars stop being innovative?

P.S. I'm *not* French or something like that...

Another safety feature - all nice and well, but...

[EkkiEkkiShiwaddle]

Allthough I'm all in favour of stashing each and every car out there with as much devices to increase safety, each time I read an article like this I still have my doubts.

The problem with all these safety features is that people feel too safe in their cars.

Going to fast? My wonderfull ABS system will bring me to a halt no matter what. Accelerating beyond my limits? Why, ASC will keep me on track. DSC will keep me on the road in those nasty corners. The new Citroen C5 has that nifty lane departure alert system, so why would I even keep my hands on the wheel, my car'll tell me when I'm flying off the road just in time!

Ok, ok, maybe I'm exagerating things a bit here, but you wouldn't believe the number of people that actually believe this stuff (or at least appear to be driving as if they believe it).

IMHO, the driver is and should always be responsible for his/her car, not some autopilot. People should be made aware of the risks of ignoring these systems more, than they should be made aware of situations they should've seen for themselves.

Know the limits of yourself. Know the limits of your car. If you go beyond either of those, no system out there now nor in future will keep you on the road.

Stuff like this scares the crap out of bikers

[peterpi]

Sure, this may be "Pretty neat technology", but not driving when you're tired is pretty neat too.

The amount of technology designed to let the car driver fall asleep is terrifying to those who actually have something to loose from an accident.

Hard times for men

[kickdown]

Damn, you wont ever convince a girl to do a blowjob in your car when she sees a camera.

Some more "unconventional" methods

[lxt]

There was an article about drivers falling asleep in a UK paper a few days back - some of the more (worrying?) ways drivers have tried to not fall asleep included a air hostess who trapped her hair in the sun roof, so when she would fall asleep the sharp pain in her head would wake her up again, and another man who attached pins around an elastic band on his wrist...

CarTrain

[wowbagger]

And building a dedicated railbed and running an AutoTrain would be even easier, far more efficent, and MUCH faster.

This is the thing that ticks me off about all of this "Intelligent Hiway" crap - we KNOW how to build trains. We KNOW how to build railbeds capable of supporting 300MPH trains. We KNOW how to build rail cars that will hold automobiles. R&D? We need no "R" - the research is done, we just need the development.

However, since we DON'T need any research, nobody wants to look at this technology - it isn't "sexy". So everybody talks about building more intelligence into the car - but of course we will need a huge quantity of money to fund research for those pesky problems like actually dealing with the one driver who's car is NOT on full automatic drive who INSISTS upon getting into that lane.

This is pretty scary.

[Puls4r]

No thanks. This scares me just like the data-recording units in some new cars. If you don't think the data is going to be stored for people to reference - 'Insurance company / police / etc', then I think you're off base.

The folks talking about automatic driving systems are also pretty unrealistic. Vision systems in use in manufacturing environments are notriously touchy and difficult to keep running, even with proper illumination and constrat control.

This will be used as a law enforcement tool. Those people who stop driving when they grow tired will continue to do so, and those that don't will disable the system and continue to drive, just like folks who refuse to wear seatbelts disable the idiot bell and light.

Again, no thanks.

The big problem with trains...

[Colin Smith]

And this is an important point. They are group vehicles.

What this means is that they have to stop at every station on the route in the offchance that someone wants to get on or off. This makes it slow, the average speed is only a fraction of their peak or rated speed. Because they are group vehicles they also have to run to a schedule which means that you have to wait for a train. Both features make journey times significantly longer than an individual vehicle would take.

Also, because the rolling stock is carrying a large group, it is heavy and relatively few of them are built which means no mass manufacture. The supporting infrastructure must also be heavy to cope with the size of the vehicle. This makes it expensive.

Ok, you've got me started now. Group vehicles also simply can't go where everyone wants to go, their ridership is only the few percentage of the population who are with in easy reach of a station. If you add more stations to increase the number of people who use it the trains have to spend more time stopped and average speed suffers further making it slower. Because they don't go exactly where you want to go you have to switch modes or lines, each time you switch you incur a journey time penalty waiting on the schedule.

Scheduled vehicles have to run whether there are people to use it or not, this kills the overall efficiency, the vehicles are heavy, accelerating them and decelerating them takes a lot of energy. During rush hour the ridership is such that it's very efficient. As soon as you get outside the rush hour period and the ridership falls so does the efficiency.

So you end up paying a lot for relatively poor performance.