When Will the Automotive Internet Arrive?

DeviceGuru writes "European researchers are developing a cooperative traffic system, known CVIS (Cooperative Vehicle-Infrastructure Systems), comprised of vehicle-, roadside-, and central infrastructure-based communications hardware and software, including vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) wireless. Among other capabilities, cars communicate with each other and with 'smart traffic signals' to smooth the flow of traffic and avoid accidents, or with 'smart traffic signs' to avoid dangerous driving conditions. The CVIS project is in the midst of undergoing field trials in Europe, and Audi has recently deployed 15 test vehicles in a similar project. The ambitious vision of intelligent transportation systems (ITS) includes goals such as reduced traffic congestion and fuel consumption, enhanced safety, and improved driver and passenger comfort. Ultimately, the developers envision a sort of Automotive Internet."

Comments

IPV6

[stavrica]

If they're smart, they'll build it out on IPV6.

(Those who consider this to be obvious should remember that the government is involved.)

Re:IPV6

[nacturation]

Too bad they didn't go with OpenBSD.

I can't wait.

[Anonymous Coward]

It is bad enough I worry about the script kiddies hacking my work computers. I can't imagine having to worry about then getting into my drive-by-wire systems which is why I'm stuck on driving old cars.

Re:I can't wait.

[couchslug]

I worked on fighters with electronic flight controls (F-16A/B/C/D) for years.

They are built far better than automobiles, which are and will remain consumer junk by comparison.

BTW, even the F-15 and F-16 engines have a stopcock mechanical throttle just in case of an auto-acceleration.

Re:I can't wait.

[Hognoxious]

Then you should also be worried about fighter aircraft which have been using fly by wire systems for quite some time now.

Do they have fly by wire systems that are connected to the big dirty outside internet full of 733t h4x0rs? Thought not.

Re:I can't wait.

[Reziac]

And what would an F-16 be like if mass-produced for the $35,000 market? and of course, if each didn't come with its own team of crack mechanics. Call it what, about 1% of the quality control an average F-16 is accustomed to??

there is a problem

[papabob]

there is a small problem with the current aproach: until "every" car gets the system installed, it's nearly useless. The protocol need to "know" that every other vehicle is going to act accordingly its specification. The false sense of security these devices can provide is very dangerous in case a car break the rules (not only by malice, just think in a malfuction like the infamous toyota) because the react time will be reduced ("The car from the back is too near, lets send a message to brake", "Ups, no response, maybe an interference, lets try again", "wow, its must be broken, lets speed up, i'll send a message to the front car to speed too", "Ups, no r...CARRIER LOST"),

Re:there is a problem

[Urkki]

there is a small problem with the current aproach: until "every" car gets the system installed, it's nearly useless.

I don't think so. I mean, even if a single car had this, and then there were roadside sensors, that single car could benefit from the sensor network. Now replace roadside sensors with just a few percent of the cars having sensors, and benefits should be pretty clear.

And once something like half of the cars would have the system, the behaviour of the other half could be predicted quite nicely within certain limits. After all, a car driving between two cars will normally (ie. until it overtakes or turns) stay between those two cars and behave very predictably.

Re:there is a problem

[Jah-Wren Ryel]

there is a small problem with the current aproach: until "every" car gets the system installed, it's nearly useless. The protocol need to "know" that every other vehicle is going to act accordingly its specification.

That's far from true. For one thing, even if every car were to have the system installed, that's no protection against bad actors.

If the developers have even half a brain they are designing the system to operate defensively rather than trustingly. That principle will limit what the system can achieve, but it also means that it will be resistant to deliberate attacks as well as accidents and non-participating vehicles. Considering that failures in the system will result in lives lost, I'd say that there is no other way to design it but defensively.

Hopefully Never

[Black Gold Alchemist]

I hope the Automotive Internet never arrives. Why? Because of three issues: privacy, security and bugs. First, this system is basically a giant handout to authoritarians and fascists world wide. One of the goals of all governments that don't care about privacy is to track every private car. They know that measure has to be phased in gradually, so we need to fight against every step of the way. Second, security is a huge issue. We know that we can never provide a %100 percent secure desktop platform - so how in the world are we going to provide a secure automotive platform? Third, bugs are going to be a huge problem - see the Toyota situation. If we have 100 million lines of code, and we have 1-2 bugs per thousand lines, we get 100-200 hundred thousand bugs in the car's software. It's surprising that we don't have more cars flying down the highway uncontrollably. I hope we have less computers in cars in the future, maybe even none if we really could. It'll be tough but it would save a lot of money and a lot of hassle.

Re:Hopefully Never

[TubeSteak]

I hope we have less computers in cars in the future, maybe even none if we really could. It'll be tough but it would save a lot of money and a lot of hassle.

Without computer control, combustion engines can't meet mileage and pollution standards.
Without computer control, electric/hybrid motors are vastly less efficient.

Unless we switch over to an entirely different engine technology, computers are here to stay.

Re:Hopefully Never

[paeanblack]

I hope we have less computers in cars in the future, maybe even none if we really could. It'll be tough but it would save a lot of money and a lot of hassle.

That depends on what your vision of the future is.

In the US, single driver commuters spend an average of 4 hours per week getting to work and back, and only a small minority rate this as a pleasurable activity. Recovering those billions of lost man-hours per year is one of the biggest benefits of an automated highway system. Furthermore, the vast majority of those cars sit idle most of the time.

An automated system has the potential to:
(a) allow those commuters to engage in productive or enjoyable activity on their way to work
(b) service multiple commuters through time-sharing
(c) store idle commuter cars on less-valuable real-estate
(d) be treated as a fleet for more efficient maintenance
(e) allow people who are not capable of driving equivalent access to transportation
(f) allow anonymous single passenger vehicle traffic.

There was a time when every elevator had a trained human operator inside, much like the modern taxi. Computerization got rid of that and allowed banks of elevators to coordinate to move more people faster and more efficiently. Frankly, turning cars into automated taxis sounds pretty cool. The biggest social hurdle is people who drive for fun.

Cost effective?

[Kohath]

At some point, it might make more sense to reduce congestion by building enough roads with enough lanes for the cars.

Re:Cost effective?

[dodobh]

Cars don't scale. Mass transit scales better.

Re:Cost effective?

[IrquiM]

There isn't enough room for more lanes everywhere.

Re:Cost effective?

[tftp]

I would suspect other places as well are finding mass transit is not the panacea the environmentalists make it out to be.

Mass transit doesn't work because there are no "masses" to transport.

In the old USSR there were buses going to and from industrial areas. So when people go to work they take those buses, and the buses are full, and the fuel economy is achieved, and it's inexpensive. Once the shifts at factories start, the buses start going less frequently, until the next batch of passengers is expected. This works because it is a highly point to point travel, from the nearest subway station (or bus transfer point) to the area's points of interest.

But a typical US city has no such points of interest that attract thousands of riders. If you map the travel routes you will see that people go from anywhere to anywhere. The fact that residential and industrial areas are randomly interleaved doesn't help either. No mass transit, short of personal computer-controlled taxis, can service such disorganized travelers. You'd need several transfers, and if you dare carrying stuff with you (like 10 grocery bags) you'd start swearing at the gods of the domain well before you get home.

Mass transit is doomed to be slow and backward way to travel just because it has to collect riders at many places. But no car owner will go to work by zigzagging through hundreds of residential streets and stopping at every minor intersection. Buses do that. Cars go straight to the nearest major road or a freeway, and proceed there, at 65 mph (minimum speed.)

There is another issue that is relevant to many US cities - safety from criminals. A car offers a pretty good protection. A bus offers zero protection, and there are many cases where buses become a crime scene. The bus driver is not going to protect you; he at best will report the crime to the overworked police, who may show up to collect your body. At worst the driver will ignore the crime - he is in the area every day, and he needs no trouble with locals. And of course once you are on the bus you can't ask the driver to skip the problematic area. In a car you never even get close to those areas.

Hopefully NEVER.

[Saeed al-Sahaf]

Hopefully NEVER.

For everything good that could come out of this, several somethings BAD will come out of it. Speed tracking for automatic tickets and insurance increases, and - NO TIN FOIL NEEDED - government tracking. The Brits will be the first to require this.

As soon as it's possible, the insurance companies will require this and jack your rates through the roof without it. As well, if your driving does not fit their statistical profile, your rates will goe up. As technology improves, if you take those right-turn-on-reds too fast, your rates will go up. Spend too much time in the "wrong" part of town? Your rates will go up.

The government will for sure figure out a way to leverage the information from this technology for some sort of tax increase.

There is no real benefit to having an Internet connected auto. Flying cars are a fantasy, road / highway technology has reached it's zenith.

Re:Don't need to have every car! Brilliant

[Brett Buck]

In fact, if you just drove a line of cars side by side along the freeway at the speed limit so that nobody could pass them, and just kept such barriers every 10 or 20 or 30 miles, then I think it would help to eliminate the incentive for everyone to act so crazy to gain 30 seconds' advantage, thereby causing congestion. I've always thought it was the lange changes and sudden maneuvers that cause the most problems in traffic.

      In a way, you are right. Idiots running side by side at the same speed causes people to figure out ways to get around them. Multiple-lane highways exist for a reason, and the *right* lane is the *slow lane* and the *left lane* is the fast lane. As near as I can tell doing as you suggest is a violation in all 50 states of the union.

          BTW, truckers passing through Kentucky on I-75 (and probably elsewhere) were protesting the different speed limit for trucks and cars by lining up side by side at the border, and running exactly the speed limit all the way across. That resulted in absolute carnage as people tried to pass on the shoulder, and lined up for miles behind them. If your proposal were implemented, I would expect a huge increase in accidents as people got around the "blocker cars".

          Traffic accidents are not caused by excessive speed to any great extent, they are caused by bad driving and discourteous driving - and your proposal is a classic example of both.

        Brett

Re:Vehicular anti-virus....

[kronosopher]

great time to invest in anti-virus

Let's invest in quality and secure software first, k?

whatcouldpossiblygowrong

[tylersoze]

If any article deserves that tag, you think it'd be this one.

Re:I can see the new billboards

[DarkKnightRadick]

So much for the "Don't text and Drive" billboards, now we'll have don't "4Chan and Drive" or "/b/ and Driving = Death you friggin B'tards"

Sheldon

Should we be warning them?!

Better

[zogger]

I think it would be loads better to push telecommuting over all this expensive rube goldberg computer controlled meatsack commuting. These are wild ass schemes that are ignoring the basic problem of physical commuting, which is the "necessity" thereof. How much is *really* necessary, and how much is just archaic holdover from the 1800s and 1900s office? Yes, people HAD to travel to the office then, because all the data was physical hard copy, all the communication was speaking directly to people or sending a snail mail or real high tech, a courier to the telegraph office. But *now*? WTF? Why are we still doing this by the millions and millions when it is all digital and can be done over wires and fiber? Why are we still insisting that people who sit in front of a computer screen have to commute daily to "the office" to do this? Aren't we past that quill pen era yet? And if they don't have to physically commute, shazzam, we don't have to waste money on these billion dollar massive corporate ego office towers either, another huge savings. Wouldn't it be cheaper to push better connectivity, run a lot more fiber, than to build more whizzbang commuter trains and computerized self aware vehicles and all that stuff just to sit in front of a computer? Isn't this the whole point of the internet in the first place, to allow communication of any type to be accomplished without having to physically move a human or a courier sack?

Re:But seriously... Big Brothermobile?

[Anonymous Coward]

Because, you know, being able to break the law without anyone noticing is more important then reducing the number of people who die on the road, yes?

If you can build such a system, you can build it so it is anonymous. Right now, from what I gather, they are mostly concerned with replacing the visual signals you already have (turn lights, brake lights, etc.) with more detailed data communications, so instead of me knowing that your car is braking, my car knows why and how hard your car breaks.

Yeah, huge privacy problem, I see that.

Frankly, half of the comments in here are posting pseudo privacy warnings, when what they are really saying is "I want to speed and drive like an asshole, and I don't like that it'll become easier for others to notice".