GM Working On Wi-Fi Direct-Equipped Cars To Detect Pedestrians and Cyclists 111
cylonlover writes "General Motors is working to expand upon its vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communication systems that allow information to be shared between vehicles and infrastructure to provide advance warning of potential road hazards, such as stalled vehicles, slippery roads, road works, intersections, stop signs and the like. The automaker is now looking to add pedestrians and cyclists to the mix using Wi-Fi Direct technology so a car can detect them in low visibility conditions before the driver does."
Why not use heat sensors? (Score:1, Insightful)
I think deer are a much bigger problem than cyclists or pedestrians.
Re: (Score:2)
Re:Why not use heat sensors? (Score:5, Funny)
You mean like this?
Phone: Beep beep!
(Pedestrian stops to look at phone)
Phone: New wireless network found: "You'll be dead in 3 seconds"
(Pedestrian gets hit)
Re: (Score:3, Informative)
But . . . Can the car warn the pedestrian that they about to get hit ?
We used to have that in cars in the seventies, it was called big V8s.
Re: (Score:2)
But . . . Can the car warn the pedestrian that they about to get hit ?
Forget pedestrians, warn the cyclists. We can call it the Jeremy Clarkson cyclist detection system.
Re: (Score:2)
Yes. If they are listening to a properly equipped iPhone or Android device, the cars computer will instruct the device to play Grandma Got Run Over By A Reindeer [youtube.com].
Re:Why not use heat sensors? (Score:5, Informative)
Why not use heat sensors?
Heat sensors are the wrong technology to use for this. Radar works much better because it can detect cold objects as well, penetrates fog/smoke, and can use the doppler effect to detect if an object is moving. Radar is what the Google Driverless Car [wikipedia.org] uses, and is what most other autonomous vehicles use as well. It is also what most automatic cruise control systems [wikipedia.org] use.
Re: (Score:3)
What makes you think radar won't detect a stationary object?
Re: (Score:2)
It will detect a stationary object, but will it detect it as a deer? Otherwise it's just another thing on the side of the road.
Re: (Score:2)
I just imagined hacking the freeway so all the cars slow down to 5 mph because of a bunch of fake deer scattered on the shoulder.
Re: (Score:2)
I would imagine, from the fact that several cars had passed the spot and shown that it didn't contain something before... And as they're communicating with each other, they now know that it's moving ;)
Re: (Score:2)
Now THAT would involve a massive leap in bandwidth - recognizing and and sharing the location of large moving objects is one thing, Distinguishing stationary roadside objects though - imagine what that would involve if you were driving through a forest...
Re: (Score:2)
But if the object is on the side of the road (not on the road), you might drive under the belief that staying on the road will be sufficient to avoid the object. If the object is a deer, that belief might be incorrect.
Re: (Score:2)
My 2007 Volvo with collision avoidance already detects stationary objects.
One of the advantages is that besides the in-cabin warning, it pre-charges the braking hydraulics so you get maximum braking power with minimum physical braking effort, giving you that fraction of a second stopping advantage in addition to the warning.
Re: (Score:2)
WRONG!
You have absolutely no clue how RADAR works. It will detect stationary objects, just as cars do when they are equipped with parking sensors.
Besides, if RADAR needed some kind of motion to work, let me give a a hint, the *car* itself is moving.
Re: (Score:1)
Heat sensors are indeed the wrong technology to use.
Some experiments have been made during winter with sub 0C temperatures (which is supposed to be the ideal condition for heat sensors), and it was just a piece of crap. It would trip when encountering shops and other well heated building, and would fail to detect some pedestrians if they are all wrapped up in good, warm modern clothes, which are designed to minimize heat radiation (duh), so you would just detect their head, unless they have a hood and they
Re:Why not use heat sensors? (Score:4, Interesting)
It should probably use a variety of sensors so that it can decide what to hit, too. Radar plus infrared might be a good way to go. Better to run over a bush than a crouching pedestrian. That combo would let you get an idea of the density (well, with THz sensors anyway) and the temperature of your potential targets...
Re: (Score:2)
It should use a 1.21 Jiggawatt fusion powered laser. Thus preventing me from hitting the bush or pedestrian by removing them from my path (and existence).
Bwa ha ha HA!
Re: (Score:2)
Why not use heat sensors?
Heat sensors are the wrong technology to use for this. Radar works much better because it can detect cold objects as well, penetrates fog/smoke, and can use the doppler effect to detect if an object is moving. Radar is what the Google Driverless Car [wikipedia.org] uses, and is what most other autonomous vehicles use as well. It is also what most automatic cruise control systems [wikipedia.org] use.
I've pondered before over how well chaff [wikipedia.org] would work to fool a collision avoidance system. If a car fitted with a radar based collision avoidance system was tailgating then a release of chaff might cause it to slam on its brakes, with ensuing hilarity.
Re: (Score:2)
I've pondered before over how well chaff would work to fool a collision avoidance system. If a car fitted with a radar based collision avoidance system was tailgating then a release of chaff might cause it to slam on its brakes, with ensuing hilarity.
It would probably work great, and you could fire it with an airbag. But you could probably also get the job done with a radar repeater, which would be reuseable and leave behind a lot less evidence.
Re: (Score:2)
For the driver, maybe. But assuming you are not an asshole who doesn't care about hitting people, there is a much bigger chance of someone getting killed in a human-car collision than it is in a human-deer one.
Bad idea (Score:5, Insightful)
If you need your car to detect obstacles for you, you're driving too fast. Because there's no way this is going to work 100% - not every pedestrian is carrying a device with Wi-Fi eneabled - so what do you do when you're relying on it and it fails?
Re: (Score:2)
Because there's no way this is going to work 100% - not every pedestrian is carrying a device with Wi-Fi eneabled - so what do you do when you're relying on it and it fails?
The enemy of good is perfect.
It isn't like people are going to start driving with their eyes closed.
Re: (Score:3)
Re:Bad idea (Score:4, Interesting)
If it is reasonably reliable - say >80% or so - it has the potential for people to take it for granted and rely on it, thereby encouraging people to drive too fast. OTOH, we already have plenty of idiots who do that without any obstacle warning gizmos, so in these situations it would certainly help. I guess to really know which effect is bigger, someone would have to do a study on how this technology changes people's behavior.
Re: (Score:2)
Re: (Score:2)
Fuck. Amen.
We should do away with all the courts and due process too. Make a lot simpler. Break a deal - Spin the wheel.
If you have an argument with another person... Thunderdome bitches. Put that on TV and the whole system pays for itself!
Re: (Score:3)
Actually there's a growing body of evidence that removing all lights/right-of-way signs in busy, accident-prone intersections can reduce the number of accidents dramatically - approaching drivers and pedestrians have no indication of what they're supposed to be doing to blindly follow, so instead they have to pay attention to what's going on and figure out their own way through. If I remember correctly total traffic throughput often increased significantly as well. So basically remove the safety / flow-co
Re: (Score:2)
Problem here is that we have an intersection just like that near where I live, and most bicycles and pedestrians avoid it like the plague. No surprise, a lack of pedestrians and cyclists drives their accident count down, but not because the intersection is "safer".
Re: (Score:2)
Well sure, if there's an "easier" intersection nearby that doesn't demand people actually pay attention then most people will go for that. But perhaps you missed where I mentioned throughput often increased, i.e. *more* people went through it in a given day, and accidents rates *still* plummeted. I'd say that's a pretty clear indication that the intersection is safer.
Re: (Score:2)
Is that "people" specifically pedestrians, or people in cars? The unsignaled intersection near me has highest auto throughput without signals, signs, or police (these have been tried), so if you are just counting "people", yes, lack of signal gives higher throughput, at least until pedestrian patterns adapt. It's easy to drive pedestrians away, not so easy to get them to come back.
You have to understand, I have one of these intersections to look at. I traverse it all the time (or find ways to avoid it, o
Re: (Score:2)
If I recall correctly several of the intersections it was tried on were did in fact have heavy pedestrian traffic. I'll admit I would be a bit intimidated by any such intersection I can imagine, and it may well be a cultural difference - I seem to remember hearing that the norm in many European cities is that if you want to cross the street you step out into it and trust that traffic will stop for you. There's also a (possibly related) factor in that if pedestrian traffic is normally heavy then drivers wi
Re: (Score:2)
If it is reasonably reliable - say >80% or so - it has the potential for people to take it for granted and rely on it, thereby encouraging people to drive too fast.
Forget the "drive too fast" argument, what about the "drive too inattentively" argument? If people drive in town and everyone has such a device (for the sake of argument) and then they drive out of town where not everyone does, they might drive inattentively because they're used to the system warning them of dangers, and then not notice someone out in the country and not drive in a way that would give someone enough space because they're used to the car telling them when they have to give someone some berth
Re: (Score:2)
Speed does not kill. Inappropriate speed kills.
Exactly. Driving fast is not a problem when the conditions allow it. But when you're going so fast that the braking distance is longer than you can see, you're going too fast pretty much by definition. Sometimes that's above the legal speed limit, sometimes below.
Of course, if you're not paying attention, the braking distance also increases because of the delayed reaction.
Re: (Score:2)
Assuming we all have the same definition of "appropriate", and whether we admit the reality of ambient human error. Given a car-ped collision, yes, speed kills. The difference between 20, 25, and 30mph is a dramatic increase in risk of death.
Re: (Score:3)
Just like cars that park themselves. IMO if you can't park your car, you have no business driving. Or people relying on ABS and Stability Control thingies...
Re: (Score:2)
Re: (Score:2)
Not what I wrote. People tend to *rely* on safeties instead of driving safely. I've got ABS, so I can drive faster. I've got Traction Control so I can just floor it. I've got Stability Control, so I can get in that curve at twice the speed. That kind of stupid thing
Re: (Score:2)
One cannot drive at any reasonable speed that will guarantee that all obstacles will be seen in time, but conservative speeds can be chosen to set the risk at a low level. For instance at night a pedestrian dressed all in black can suddenly appear too at a distance too close for the driver to be able to stop. If the system is detecting problems before they are seen a prudent driver might decide to slow down to slow down.
Some drivers will probably think that the system (like anti skid brakes) will provide
Re: (Score:1)
Because there's no way this is going to work 100%
The enemy of good is perfect.
Especially on Slashdot. Insisting that anything that is imperfect is therefore worthless is a common trait of Asperger's Syndrome [wikipedia.org].
Re: (Score:2)
Your wiki link does nothing to backup your claim. Where did you pull it from?
Re: (Score:2)
Of course they're not going to START driving with their eyes closed. They already DO!
Re: (Score:2)
If they don't have wi-fi, then they are WHAP enabled.
I think, 'tho, the problem would be trying to enter the password to connect to their network first.
Re: (Score:3, Insightful)
Every citizen will have a Wi-Fi enabled GPS tracker 'installed' in them .... for their own safety!
If you've got nothing to hide .......
Re: (Score:2)
Driver: "Why do you call them crunchies?"
Vehicle: "From the sound they make when they go under my tires, sir."
Re: (Score:2)
I live in a suburb where most corners and intersections are obstructed by high hedges, trees or parking cars. You literally can't see if another vehicle is coming until you are right in front of it. While I know this technology won't give 100% safety, widespread use of it could reduce the risks significantly.
Re: (Score:2)
Re: (Score:2)
Till pedestrians start pushing shopping carts full of concrete blocks into intersections before they step into them.
more stuff (Score:1)
to get hacked, especially knowing how good american companies are at making security systems....
Great (Score:2)
Drivers already don't look where they're going when using a GPS, now they won't lookout for dangers or other users of the road because the car will tell them.
Re: (Score:2)
Drivers already don't look where they're going when using a GPS, now they won't lookout for dangers or other users of the road because the car will tell them.
Ford is selling a car with "parking assist" and advertised as "almost self parking" yet people in a brand new Focus still cant park.
Self-Driving cars (Score:5, Interesting)
As a motorcyclist, hell, I'd trust a TI-85 with a camera to steer, over the uncomfortably large percentage of SUV drivers that occasionally interrupt their texting sessions by glancing up at the road. Anything that improves the technology to prevent careless accidents is good in my book, and I would think the most beneficial application would be in respects to the self-driven cars, like the ones Google is developing [wikipedia.org], no?
Re: (Score:2)
I'm missing something, I believe. Help me out.
over the uncomfortably large percentage of SUV drivers that occasionally interrupt their texting sessions by glancing up at the road.
Are you saying that it would be better if the operators of said 2 ton high speed machinery driven within inches of your handlebars DID NOT EVER look up? Are you assuming that these folks are keyboarding vehicular control commands from their phones? Has someone ported EMACS to the iPhone?
Re:Self-Driving cars (Score:4, Interesting)
As a fellow rider I concur wholeheartedly.
However one question does occur to me... when self-driving cars become the norm, at some point won't it be obvious that regular cars are a hazard and the sale of non-autonomous vehicles is banned? Where does that leave motorcycle enthusiats?
Then again, daytime running lights STILL aren't required on new cars sold in the U.S. (and those would definitely save lives) so I'll probably be an old man before they even start thinking about it.
Re: (Score:2)
Re: (Score:1)
Re: (Score:2)
I think there's a good argument to be made that small vehicles are considerably safer (for others) than large ones in a collision, which would have to be factored in to risk assessment. There's also the fact that driving an inherently unstable two-wheeler requires a much higher level of driver alertness just to remain upright. And the total absence of a crash cage tends to get the whole survival instinct thing incentivizing obstacle avoidance to a much greater degree.
Hopefully the (still slowly) growing tr
Hack (Score:5, Funny)
Prediction (Score:1, Funny)
Since the Government Motors company doesn't really have to care about money anymore, it can waste resources on such things rather than working on better cars, they'll be doing these types of gimmicks, then, when they have 'something', the government will pass a LAW that nobody can have a car without such a system installed in it, the rest of the manufacturers will be forced to pay for licensing of these patents and then the prices for the cars will go up and choices for the cars will be diminished in USA.
Fi
Re: (Score:2)
As it turns out, the vast majority of your prediction already happened. With a fair amount of it implemented by the prior republican president and the long time republican congress. Not that that has anything to do with anything, but I figured I'd try to counter some of the crazy.
Chances are pretty good you're already in a facial recognition database and you've been recorded 100 times by cameras every day.
Re: (Score:1)
Yeah, almost all presidents, almost all Congressmen, almost all Senators, almost all judges, they are almost all complicit in this power grab. There are exceptions, but they are exceptions to the rule, they are not the rule.
When another POTUS comes around he wants to do 'great things', well actually he shouldn't be doing any 'great things', he should be upholding the law, that's all.
Re: (Score:2)
You forgot the part about how the RIAA/MPAA will be monitoring the sounds in your car to make sure that you are not listening to songs or movies that you haven't purchased the CD/DVD/download for according to their database. If they detect an unauthorized song or movie, that's when the police drone launches its missiles.
If you're lucky.
Re: (Score:1)
You aren't going to be that lucky, instead of a missile, the drone will launch a lawyer.
Here's the funny thing about all of this... (Score:4, Interesting)
The funny thing is that almost none of this will matter at all. Over the last 4 decades we've employed a variety of engineering improvements like air bags, anti lock brakes, better tires and suspensions, backup cameras, crush zones and so forth. This reduced the accident and death rates through around 1990-1995. Since 1990, those rates have remained almost exactly the same, year on year.
This means a couple of things. One is that cell phones had no effect on accident rates, because they've remained the same from 1990-2010. It also means that the crusade on drunk driving had no results as far as reported accidents. It also means that this system will have no beneficial effect until the driver is removed from the equation and the technology is perfected to a level where its significantly better than the driver it replaces.
Driving is boring to most people that have been doing it for a while, so they seek distractions from the boredom. Doesn't matter what distraction or tool you add to or remove from the equation, we'll fix our boredom somehow. In the 70's when I almost got run over by someone in a parking lot, you couldn't scream "Put down the %$@#ing phone!", it was "Damn woman driver!" or some such.
Further most people are simply unaware of the simplest rules of the road like right of way, proper turning, safe following distance and so forth.
So if you don't know what you're really supposed to be doing and you're actively looking for escape from the primary activity, adding some iffy technology that can't do much better than 70-80% in effectiveness will simply further reduce our interest in paying attention to driving.
And $10 says we'll get the same exact accident rate if and when this technology is deployed.
Re:Here's the funny thing about all of this... (Score:4, Insightful)
One is that cell phones had no effect on accident rates, because they've remained the same from 1990-2010. It also means that the crusade on drunk driving had no results as far as reported accidents.
Does the concept of multiple regression mean anything to you? Maybe the cell-phone effect cancelled the decrease-in-drunk-driving effect?
occupant deaths went down, ped/bike deaths up (Score:5, Interesting)
Over the last 4 decades we've employed a variety of engineering improvements like air bags, anti lock brakes, better tires and suspensions, backup cameras, crush zones and so forth. This reduced the accident and death rates through around 1990-1995. Since 1990, those rates have remained almost exactly the same, year on year.
Meanwhile, pedestrian and cyclist deaths have gone up because US road safety consists of "make crashes as survivable as we can for the people in the cars, because we've felt they are inevitable." As a result, the death rates for peds and cyclists is 5-10x that of countries where there are vulnerable user laws. Basically: if you hit a pedestrian or cyclist - you have to prove it was their fault, and if you can't, YOU are assumed at fault. Not the other way around, where we assume it was the fault of the pedestrian or cyclist. Such an injury or death is also a criminal matter.
Re: (Score:2)
The choice seems pretty obvious then...hang up the sneakers and the bike and get into a car where you'll be safe.
Re: (Score:2)
Which, unfortunately, actually raises your chance of an early death, because lack of exercise is a massive killer. Danish study, non-bicycle commuters, 39% higher mortality rate, corrected for risk factors. Variations on this show up again and again -- excess sitting kills, elderly walking extra adds years, etc. We're doing it all wrong, if the answer is more use of cars.
Re: (Score:2)
See, I just bought a house with four floors and I have stuff all over the house so I put on a lot of miles and steps just walking around the house. Also bought in a ridiculously hilly area (as one could imagine since the town name has the word 'hills' in it). According to my fitbit, I average about 15,000 steps and 20 stories worth of steps per day, and I'm in excellent condition without using a bike or walking in traffic.
ugh (Score:2)
I dont carry any wifi devices, so when I get ran over by some troddle dick its now my fault cause he wasnt paying attention
Spoofing? RIAA? Targeting? SETI? (Score:1)
How long until someone tapes a "Wi-Fi Direct-enabled smartphone" to someone's car and the app is set to go off randomly? Or just puts a transmitter in the middle of the street and sets it to go off randomly?
How long until the RIAA jumps on the words "peer to peer" and that "music files or contact information could also be securely transferred from the home computer to a vehicle’s infotainment or navigation system" via Wi-Fi Direct devices?
How long until a deranged geek realizes that anyone running a
Re: (Score:2)
To put things into proportion, how long till some drunk or texting clot wipes out some unlucky pedestrian?
Less than 24 hours, almost guaranteed (3000+ peds per year killed by car crashes, or 8 per day).
My message to other drivers . . . (Score:2)
The route I am taking to work has:
stalled vehicles, slippery roads, road works,
earthquakes, zombie attacks, and a paparazzi drag race with Justin Bieber and Lindsay Lohan.
You should avoid driving on my route to work.
Such a system is just begging to be Black Hatted.
Re: (Score:1)
Just bought a '71 Ford F250 (Score:1)
And boy am I pleased. This vehicle will last beyond my life expectancy and there is nothing nanny about it, I tell you what.
Re: (Score:1)
Re: (Score:1)
So you're 55-60?
Cop recognition (Score:2)
I've wondered if crowdsourcing might eventually result in auto-recognition of cop cars and a warning sent out to other drivers. Maybe even a system where the locations of all police vehicles are broadcast in real time.
Re: (Score:2)
That will stay legal for about 5 seconds.
the evil that men do (Score:2)
I wish I could say "sorry" or "thanks" (Score:2)
The basic reason for that is that, there is no way for someone to apologize for a minor infraction. If I cut some one off unintentionall
Re: (Score:2)
I'm sure if you beeped out "sorry" in Morse code, they would understand.
So to apologize, just go:
"bip-bip-bip beeep-beeep-beeep bip-beeep-bip bip-beeep-bip beeep-bip-beeep-beeep"
Re: (Score:2)
That would be a nifty idea although it may just be another distraction. A change in attitude might be better... maybe coupled with a big red light on the top of the car. If the light isn't lit then a "oops, sorry, my bad" is implied, and the driver would need to illuminate the light to reflect a "haha I meant to cut you off just then!" intention.
A few years ago I was stopped at a give way sign and someone rear-ended me. It didn't really upset me that much at all... she obviously wasn't paying attention but
UAC + WiFi Direct Cars (Score:2)