For those who want to record your driving data for personal use, as opposed to real-time dissemination, it's amazingly inexpensive.
Last week, I bought an ODB2 Bluetooth adapter on eBay for about $12. You plug it into your car's diagnostics port, below the steering wheel. (It's the same port that the DMV plugs into to check emissions stuff, or that Progressive Insurance uses to give a safe-driver discount.)
With the adapter plugged in, I synced it with Torque Lite, a free Android app on my phone. (There's a $5 paid version that offers more features, but the Lite version offers enough for me.)
In the app, I could see (and record) my speed, estimated MPG, fuel level, acceleration, ambient air temperature, engine coolant temperature, engine load, engine RPMs, fuel pressure, throttle position, and more.
I picked up the same combo last year. It really is a nifty setup. My problem is that I rarely use it. My driving patterns are irregular enough that I can't safely leave the device plugged in, drawing power. The ODB2 port in my car is very inconvenient to get to, so most days the device just sits in the glove box gathering dust.
The lights on my OBD2 adaptor come on when I plug it in, no key in the ignition. I can only assume for how cheap and dumb the device is that it has the bluetooth on all the time.
Its not needed, its a technology already eclipsed by what manufacturers and even Google are doing. Self driving cars rely on various sensors; RADAR and laser to name two; to protect from collisions and the like. They do all this without communicating with the vehicles around them.
Simply put, smart cars don't need to input of other vehicles to know how to perform safely, in fact not relying on other cars makes them safer as they cannot be spoofed nor act on erroneous information.
Then of course there is that big privacy issue that this proposal tries to skip over.
Radar and lidar aren't foolproof. Radar's easy to jam and needs complicated processing to differentiate between clutter and things that might cause issues. Lidar breaks down in fog and dust and has fairly limited range.
There's a reason air traffic control have moved to active ADS-B transmitters rather than simply using ground based radar to track aircraft - reliability. There is definitely safety issues from spoofing of data, but if those issues can be dealt with (perhaps using the certificate based system that's discussed in the article), reliability compared to car-mounted radar or lidar is significantly greater.
The lidar mounted on Google's autonomous cars is a long way from being integrated into production vehicles. The V2V proposal draws on existing car sensors - all that's needed is the addition of a small chip. V2V is a manufacturer proposal, and it's almost ready for production. It's a cheap and easy method of vastly increasing car safety. Road traffic accidents kill and injure a huge number of people. We should embrace anything that makes them significantly safer.
While I agree that they are not foolproof you do not use them separately. A combination of active and passive systems is used by the car to understand its surroundings.
I would never compare the electronics used in aviation to anything employed in cars, the level of trust in aviation and the regulations behind it are legion. I seriously doubt that manufacturers would want to be under similarly complex and expensive regulation
The V2V proposal draws on existing car sensors
all that's needed is the addition of a small chip.
A smart response to a vehicle ahead suddenly braking requires both vehicles to know what road they're on, how many lanes that road has, and which lane they're in.
Maybe your car already has all that high-tech equipment, but my Honda CG125 sure doesn't!
I agree, and would also like to point out that these are not mutually exclusive. Lasers can help detect pedestrians that run in front of cars, while intercar communication can send information about cars "intentions" and additionally could send warnings back to other cars about obstacles they may not be able to detect for a ways out.
I'm torn on the subject of car tracking. Firstly it's not like it doesn't happen already, it's just that the authorities don't have access to the data. Secondly, drivers in a lot of countries are simply morons (read: you meet several when driving for 30 minutes), though I can't say I've met a lot of them in US.
On the other hand, there are the privacy and security (and possibly safety) issues...
But personally, I wouldn't buy any car that has any outside communication, possibly except for locks, heh.
I can't wait for this to happen. I'll take the transmitter out of my car, tape it on my body, and walk on the streets downtown during the rush hour. No car will be able to hit me :)
Why do I get the impression the security for this will be treated very superficially, and they'll only try to fix it as an afterthought later after many reports of hacking into such systems, just like for everything else? This worries me quite a bit.
Also it's interesting that this comes out almost at the same time with EU's plan to let the police remotely stop cars' engines:
For those who's not familiar with V2V standard, or think this is useless, here's some more context.
The whole standard stack is called Wireless Access in Vehicular Environments (WAVE) / Dedicated Short Range Communication (DSRC). Short range is as in being shorter than cellular networks, which is easily over hundreds of meters.
Physical Layer and Link Layer of WAVE/DSRC are defined in IEEE 802.11p. It's so much like other protocols in IEEE 802.11 suite that it's only an 802.11 amendment and is now integrated into IEEE 802.11 standard. In US regulatory domain, 802.11p uses 5.855 GHz - 5.925 GHz, which is slightly higher (and not overlapped with) other 5 GHz 802.11 protocols like 802.11 a/n/ac, so that V2V won't be interfered by WiFi devices. Above Link Layer is not conventional. It's designed to support TCP/IP on top of that, but just for non-safety messages. Safety messages are supposed to use WAVE Short Message Protocol (WSMP), which replaces IP Layer and Transport Layer in TCP/IP suite. WSMP is believed to be more efficient in V2V context.
A promising application of WAVE/DSRC is Cooperative Adaptive Cruise Control (CACC) and Emergency Braking. Consider following scenario:
---------------------------------
A B C D E
---------------------------------
where ABCDE are five vehicles on highway. In ACC, each vehicle uses radar, infrared, etc. to detect distance between the vehicle in front of it, and adjust speed. Theoretically this may reduce safety distance because machine reacts faster than human being and in case of emergency the vehicle can brake by itself. However, it still cost time to detect the change of distance, to calculate, and to brake. So if A encounters collision, B would brake, then C detects distance change, and brake, then D brakes. When it comes to E, accumulated time is much longer. This means you can't really reduce distance that much if you want to rely on automatic emergency brake. With V2V enabled, each of the vehicle keeps broadcasting vehicle broadcasting so they all know each other's condition. If A encounters a collision or brakes very hard, E would know almost at same time when B knows. This means E can react much faster, so vehicles can be much closer to each other. This is very useful for trucks considering the aerodynamics. Also, with vehicles closer to each other, you can put more vehicles in the same high way.
Other applications include interaction collision warning, travel time prediction, and integration with pedestrian safety system etc. The key thing is V2V is localized, distributed, and thus real-time and more responsive than cellular connection. It's definitely gonna enable more exciting technology advances in automotive industry.
I doubt vehicles are special enough to require their own replacement for TCP/IP. Last time I heard that kind of talk was with WAP/WML – anyone remember those?
TCP/IP is good enough, and making your own protocol on that level is never worth the huge hassle for everyone involved. Unless your aim is to obfuscate and erect barriers to entry.
Upvoted. I agree that it's annoying and confusing. I don't like it either. But that's the standard that IEEE, ITS and USDOT are supporting and car/truck companies are working on that. Unlike WiFi routers, vehicular networking requires everybody to use the same standard in order for it to work. I'm not sure how much space there is to make an effort to build an alternative.
They have their reasons for replacing TCP/IP, among which one is to reduce IP overhead in vehicular environment. Since most of time communication is temporary and doesn't require routing, IP layer is not necessary. They want to reduce the IP header overhead for each packet. I doubt how much it improves reliability though, given the fact that people have to implement another whole stack other than just using mature TCP/IP stack.
Supplementing unreliable warning systems - they will be unreliable given that you'd have to wait for older cars to be cycled out of service - for good driving habits seems like an obvious recipe for disaster.
Even among good drivers, I doubt many people really want another device competing for their attention when they're trying to make a decision.
#
So, I suspect somewhat more sinister motives. It would be a lot simpler for law enforcement if they could just stick up small stations to monitor everyone within an area, rather than having to put speed cameras and so on there, and it would allow other behaviours than speed to be automatically analysed.
Last week, I bought an ODB2 Bluetooth adapter on eBay for about $12. You plug it into your car's diagnostics port, below the steering wheel. (It's the same port that the DMV plugs into to check emissions stuff, or that Progressive Insurance uses to give a safe-driver discount.)
With the adapter plugged in, I synced it with Torque Lite, a free Android app on my phone. (There's a $5 paid version that offers more features, but the Lite version offers enough for me.)
In the app, I could see (and record) my speed, estimated MPG, fuel level, acceleration, ambient air temperature, engine coolant temperature, engine load, engine RPMs, fuel pressure, throttle position, and more.