6 ms·
This is all fascinating stuff, thanks. Of course, in the future when all cars are connected (even if they're not self-driving), combining all this sensor data
by 3chelon 10y ago
This is all fascinating stuff, thanks.
Of course, in the future when all cars are connected (even if they're not self-driving), combining all this sensor data and sending it to the vehicles behind you would save countless lives. The number of times you hear about pile-ups in fog and heavy rain - that could all be avoided by cars talking to each other.
But I'll bet there's still no open (or even closed) standard on inter-manufacturer comms yet?
- spydum 10y agoand all the script kiddies will find it great fun to send out those "collision" signals on a busy roadway just to screw with people. the futures gonna be great!
- arethuza 10y agoThat does raise interesting security problems though - my car receives a message saying to stop as quickly as possible, how does it decide whether this is a valid message or someone hacking cars for fun?
- Splines 10y agoI don't have an answer for you, but maybe it stems from some sort of idea of identity? Maybe the car sends a message authenticated with your license ID. Cars are (hand waves) able to authenticate the validity of a message. If you send a message saying you were in a collision, that comes with the side effects of saying you were in a collision - emergency contacts and police may be notified, maybe something pings cars registered to you to see if it was actually in an accident based on sensor data. Your insurance rates may go up. If it turns out you were lying, maybe your car gets flagged for repairs next time you try to re-register your tags. Or maybe you are flagged for trolling and you get the "boy who cried wolf" effect.
- 3chelon 10y agoSure, the actual vehicle systems had better work and be properly authenticated, and that could be achieved in numerous ways... but none of this stops a hacker war-driving with a laptop, or sitting on an overhead bridge, or on the other side of the world. There is no doubt that intelligent highways would be a massive target for terrorist and cyber attacks. However, one thing in our favour is the fact there is so little technology already in place, since they won't be burdened with decades' worth of insecure legacy systems: at least they'll be able to start out building infrastructure in full knowledge that it better be secure.
- athenot 10y agoI'm tempted to qualify the second point as optimist. US voting machines didn't have to deal with legacy technology and still turned pretty bad from an infosec POV.
- mannykannot 10y agoExactly, and it is not just this one case: serious, avoidable security cock-ups have happened every time a new class of device has been put on the internet, with the IoT DDOS debacle being perhaps the latest example.
- abfan1127 10y agonot burdened by legacy, also means not burdened from years of lessons learned.
- 3chelon 10y agoSadly, the lessons learned in legacy systems do not necessarily address the problems we have now. Older systems were typically combating noise, signal degradation, power consumption, processor speeds, poor user interface technology, etc. Often they were extremely proprietary and if they included any kind of security it was commonly by obscurity, and even if not, the crypto used would be easily crackable by a modern script kiddie packing Aircrack or similar. The lesson we do need to learn is future-proofing: whatever infrastructure we install on the millions of miles of public highway will need to be serviceable for decades.
- gm-conspiracy 10y agoSo, a malicious actor could spoof the vehicle of somebody they dislike, and cause their insurance to go up, or pave the path for a later incident, where when police/emergency are contacted, the person is ignored?
- 3chelon 10y agoI agree, that is probably going to be one of the most difficult problems to solve. However, as we dive headlong into self-driving tech, we are going to have to find an answer since the tech will be so much easier to implement if the vehicles communicate.
- bradleybuda 10y agoPKI, under the current proposals: http://www.nhtsa.gov/staticfiles/rulemaking/pdf/V2V/Readiness-of-V2V-Technology-for-Application-812014.pdf http://www.nhtsa.gov/staticfiles/rulemaking/pdf/V2V/Readines... (PDF link)
- semi-extrinsic 10y agoCouldn't this be solved better by some form of web of trust (WoT)? Say your car A trusts other cars B, C ... Y to varying degree in a WoT. It meets a brand new car Z that is not yet trusted by any cars A-Y. Car A receives a message from Z, to which it assigns 0 credence points, meaning this data is not taken into account in any decisions. If car A then makes an observation confirming the message, car A now trusts Z by +1 point, and lets other cars know this. (I'm unsure how to make aggregation/distribution of trust work smoothly, but I believe the random mixing of traffic patterns will work in the system's favor.) So gradually car Z becomes trusted, up to some maximum, e.g. 100. The only worry then is a malicious actor who mostly sends true messages, but that's hard to distinguish from a car with a sometimes faulty sensor. Those problems could be handled the same way, namely aggressively penalising a car for sending incorrect data (e.g. by -30), such that it's quickly distrusted, together with a prominent warning in the failing car when it goes below 0 trust that it needs maintenance (after which trust would be reset to 0). Possibly also increased insurance premium if your car remains with a trust below 0.
- sandworm101 10y agoThe same way you determine whether that stop sign is legitimate or not: you don't. You stop regardless of the possibility that the sign was placed there by potential carjackers. Until the hacking risk is demonstrated to be statistically more dangerous than ignoring potential warnings, all warnings will be trusted. Regardless of current proposals, any manufacture that ignores reasonable electronic warnings (ie STOP accident ahead) on technicalities (ie an expired cert) risks serious liability. This touches on the much more likely scenario whereby the car's internal data doesn't mesh with the data from it's sensors. If the sign says "one way" it is a one way street. What google maps says doesn't matter. Current warning signs trump previously recorded data. "But google maps said I could do 70!" will be no defense.
- hyder_m29 10y agoThat's a great way to rob people on an otherwise deserted highway.
- komali2 10y agoAs always Science Fiction swoops in - in the "Culture" series, AI have hardcoded (as in, on their physical hardware) automatic code to shut down external communication in the event external communication is attempting to takeover the AI. I mean your question is pretty much "what would that automatic code look like," and I have no idea. I'm just getting pretty excited about all this sweet stuff happening with AI, automatic cars, hijacking of the human nervous system. Hngg.
- semi-extrinsic 10y ago> AI have hardcoded (as in, on their physical hardware) automatic code to shut down external communication in the event external communication is attempting to takeover the AI This sounds very much like you're describing an antivirus program, which we all know works perfectly well /s
- JustSomeNobody 10y ago> ... when all cars are connected ... You're so funny. This is a world of proprietary technology.
- 3chelon 10y agoWhat can I say? I'm an idealist :)
- ryandrake 10y agoExactly. At best we'll have 15 different standards, and Fords only talk to Fords, Toyotas only talk to Toyotas, and they all speak different incompatible protocols.
- iamatworknow 10y agoJust think of all the integration programming jobs that will be preserved! All of the pointless conference calls and GoToMeetings scheduled weeks in advance between developers and project managers and teams on the other side of the world because someone made a typo in an API specification that could have easily been cleared up in a two-second email.
- ryandrake 10y agoGood times ahead! The hardware stacks will all likely be different too, so gravy train for all the HW and systems folks too. dozens of companies all re-inventing the same thing is why tech still has full employment.
- Gracana 10y agoActually, the hardware guys already have it solved. We already know how to do bus arbitration, and cars are like little buses. CANbus could work, and it's already used in the automotive world. It's nice because the higher-priority bus driver is not delayed. Unfortunately, the lower priority one gets run over. Might have to work on that.
- 10y ago
- mseebach 10y agoYou could mostly prevent pile-ups, but pile-ups generally don't kill people. Also, you don't need connected cars to avoid pile-ups, you just need to keep your distance and your eyes (sensors) on the car in front of you. Both of which works pretty well with current (unconnected/unilateral) technology. None of the causes in listed in this article require interconnection to avoid: http://uk.businessinsider.com/the-cause-of-the-most-fatal-car-crashes-2015-5 http://uk.businessinsider.com/the-cause-of-the-most-fatal-ca...
- ctrl-j 10y agoMost of the pile-ups that happen around where I grew up were due to thick fog. Where by the time you see the wreckage, it's too late. Now, that can be solved by people slowing down due to conditions. But how slow is slow enough? Often it's slow enough that the person behind you is gonna wreck into you. Maybe pile-ups aren't deadly, but wouldn't it be better if we had tech that helped prevent them? Or do maimings not matter?
- mseebach 10y agoSlow enough is slow enough that you can see an obstruction soon enough to stop before driving into it. Either way, interconnections won't help you -- what if the car is crashed hard enough (or long enough for the battery to run down) that it's beacon is off/broadcasting wrong information? What if there's fallen tree or a pedestrian on the road? Generally, sensors (radar) can see through fog, and that's the most help you're going to get in a situation like this. Yes, maimings do matter. But they don't count towards "save [ing] countless lives".
- ctrl-j 10y ago> what if the car is crashed hard enough (or long enough for the battery to run down) that it's beacon is off/broadcasting wrong information? Doesn't that mean we're just back at the situation we're in today? The worse situation? > What if there's fallen tree or a pedestrian on the road? Also, not the situation at hand. Talking about the case where a car has crashed and is letting the cars behind it know that it has done so. It's like electronic flare deployment (or hazard lights.) Don't you think that a lot of lives and livelihoods could be saved if we had such functionality?