Toyota's new accident avoidance system will take control of the car.

Do you think they can design in the ability for it to look for every possible outcome or circumstance?
For the system to be viable it would have to keep track of all those things for one, for two programming some safeguards like not going off a mountain or into other objects is not exactly rocket science. The bigger issue is varying surfaces, as someone above posted, what if the car decides that going right is fine but on the right is gravel and you are flying onto it at 50 while under full braking? Doesn't seem like it would end very well and I'm not sure how complicated surface recognition would be.
 
Do you think they can design in the ability for it to look for every possible outcome or circumstance?

Do you think most (or any for that matter) drivers can look and respond appropriately to every possible outcome or circumstance?
 
Do you think most (or any for that matter) drivers can look and respond appropriately to every possible outcome or circumstance?
The problem is that it is one thing to fail because of your own mistake a completely different issue is when a computer makes a wrong decision and causes an issue. To make it worse what if I had the right idea and would have been able to escape the accident or minimize damage but the computer took control away from me and made a mistake.
 
There are always going to be unavoidable accidents, like getting rear ended at a stoplight, no computer can do anything about that.

I don't think any computer can see the road as far ahead as a properly alert driver is supposed to be looking to avoid getting in a situation where you have to make a quick reaction in the first place, either. Most accidents are avoidable by being alert and attentive enough.


I'm thinking something like this could even backfire, leading to an increase in accidents. If people get to thinking their car will get them out of any trouble and keep them safe, the more careless they will drive.
 
properly alert driver
That's the main problem isn't it? :p
getting rear ended at a stoplight, no computer can do anything about that.
Computer in the car behind you that would have rear ended you but didn't because it's computer decided to brake earlier/harder than the driver could prevent that accident :)
^Just playing devil's advocate here really.

I don't like the idea of giving control to the car, until all cars are self driving and networked I will not trust that system.
 
Last edited:
There are always going to be unavoidable accidents, like getting rear ended at a stoplight, no computer can do anything about that.
Sure it can, it needs to tell the car behind you to brake in time for it to stop. With networked cars, that's easy. Networked cars is no doubt the future (after we have settled on a standard so every car talks the same language). Already works on planes (altough not automatic, and only as long as pilots obey their computer overlords, and not stupid humans (Bashkirian_Airlines_Flight_2937) and planes after all go pretty fast. Air traffic safety have gotten further than cars (safest way to travel), but adaptations of the same technology should trickle down to the automotive world (cost issue).

Already have auto braking by cameras which prevents a portion of rear endings and decreases the forces involved even if there is a collission. Add network capabilities and Bill Gates is your uncle.
 
Last edited:
You also introduce the ability to remotely hack a car with potentially deadly results.
 
Air traffic safety have gotten further than cars (safest way to travel)
There are also quite a bit fewer planes in the skies and they are spread much further than cars (outside of landing/take off) and are also flown by people who went through much more rigorous and advanced training. Not to mention that most commercial pilots are former military (only real way to get enough flying hours).
You also introduce the ability to remotely hack a car with potentially deadly results.
You mean how everyone and their mother hacks cell phones through the air with no problem? No wait they do not :p We have ways of making sure radio transmissions are secure one of those is broadcasting on a licensed band that requires specialized equipment to use.

BTW you can remote hack a car with potentially deadly results right now. Get a .22 with a scope (available at sporting goods stores), get somewhere with a good view of a major highway where people go fairly fast and start shooting out tires.
 
Last edited:
There are also quite a bit fewer planes in the skies and they are spread much further than cars (outside of landing/take off) and are also flown by people who went through much more rigorous and advanced training. Not to mention that most commercial pilots are former military (only real way to get enough flying hours).

You mean how everyone and their mother hacks cell phones through the air with no problem? No wait they do not :p We have ways of making sure radio transmissions are secure one of those is broadcasting on a licensed band that requires specialized equipment to use.

BTW you can remote hack a car with potentially deadly results right now. Get a .22 with a scope (available at sporting goods stores), get somewhere with a good view of a major highway where people go fairly fast and start shooting out tires.

You're thinking on the wrong scale - you don't need access to jam a signal. Not a big deal to cut off someone's cell phone, but cut off communications to all vehicles speeding down a highway.....
 
You're thinking on the wrong scale - you don't need access to jam a signal. Not a big deal to cut off someone's cell phone, but cut off communications to all vehicles speeding down a highway.....
Easy enough to add a safeguard that will have all the affected vehicles gently brake to stop, they would still have to have sensors and radars (pedestrians, rocks, and wildlife don't come with WAN capability :p)

Sure a cell not working is [usually] not as critical as a car but at the same time cell phone networks are prone to the same kind of jamming/hacking possibilities as a theoretical network for v2v communication and we have yet to see any kind of attacks on those so it stands to reason that same would apply to a V2V network. Also cars wouldn't necessarily require any kind of central infrastructure (though there may be a certain benefit to a central traffic control system) and could easily use a P2P (or rather V2V :p) style network. The communications can be short range (only matters what vehicles in your immediate vicinity are doing really as any changes will be relayed vehicle-to-vehicle anyway) and encrypted, add to it a dedicated band and you are making it a very difficult network to crack and/or jam. Also don't forget the fact that cars in question are likely moving at 50-60mph so you would need to cover a very large area AND you would need to cause some sort of a disruption at the same time. Otherwise the cars will simply go through the black out area at constant speed and nothing will happen.

Besides a large scale traffic disruption like you are talking about is more than possible now w/o any need for a V2V communication network. Simply set up a power EMP near a major highway, that means that anything with electronic fuel injection (which is something like 99.9% of cars on the road) will stall out, which means no power assists for brakes or steering (especially if you have electric steering assist like some Mazdas, likely yours as well). Add fairly high speed and density of traffic (say tail end/beginning of rush hour with many cars on the road but not to the point where traffic is slow moving yet) and you got a recipe for a major disaster. This sounds about as plausible as a powerful enough radio jammer to affect a large number of fast moving vehicles.
 
That powerful of an EMP device is not exactly easy to procure, might as well be afraid of a tank rolling down the highway (yes yes don't post that video). But you make a good point of peer-to-peer networking. That does change things, I was thinking of central servers.
 
That powerful of an EMP device is not exactly easy to procure, might as well be afraid of a tank rolling down the highway (yes yes don't post that video).
Wasn't going to post it or even mention it, it's one of those possible but unlikely things.
But you make a good point of peer-to-peer networking. That does change things, I was thinking of central servers.
It would likely move to central eventually there would be some advantages to that such as ability to reroute traffic based on road conditions and synchronize lights with vehicles to allow for the greatest efficiency of the network.
 
The lights could be listening to the traffic around and adjust themselves in the swarm of communicating lights all over the city. No need for central entities :lol: just a swarm of smart lights.
 
The lights could be listening to the traffic around and adjust themselves in the swarm of communicating lights all over the city. No need for central entities :lol: just a swarm of smart lights.
A centralized system would be better to have a big picture view of the traffic and reroute as necessary. I'm talking telling cars to go different routes not just lights switching around.
 
I wouldn't really want a centralized system... it would probably monitor speeding and there are some ridiculous speed limits around here (ex: a 4 lane divided road, speed limit 25 mph. Are you fucking kidding me town? that is PAINFUL).
 
The swarm would take care of it. Think IP, just with cars.
I get your idea, its interesting but what about "packet collisions" :p
I wouldn't really want a centralized system... it would probably monitor speeding and there are some ridiculous speed limits around here (ex: a 4 lane divided road, speed limit 25 mph. Are you fucking kidding me town? that is PAINFUL).
This assumes self driving cars so you wouldn't be speeding because you wouldn't be driving. Also like I said earlier speed limits are likely to either be raised or abolished alltogether as there would be 0 revenue from speeding offenses and thats really the main reason for the limits. It would be more likely that cars would slow down in town to normal 30-40mph and go 100+ on highways. They would also likely be electric by then so environmental concerns would also be non-existant.
 
Last edited:
I think this is an awful idea.

For years and years the general motivation behind car development seems to be "let the car do it", rather than "teach the driver how to do it". It is such an awful mentality and is the first step to cars being physically driven being phased out and then banned due to being "un-safe".

Just look how well Volvo's car stoppage system went when at its launch when the car drove straight into the barrier. If anything it shows that no matter how good something is it won't be fool-proof. Not only that but how long will it be before people drive more carelessly because "don't worry, the car can prevent crashes itself".
Not to mention it will leave cars open to, in theory, being hacked. Hell even file corruption in the computer system could fuck the whole car.

The best and safest drivers I know are ones who drive older cars. Once you realise you have 40 year old tech in you braking and suspension system you drive a whole lot more carefully. Especially with the knowledge that if you crash your remains will be buried in a jam jar...
Safety devices give IDIOTS and bad drivers false confidence, this system is aimed at IDIOTS and bad drivers.

I despise the very idea of this sort of system...
 
Just look how well Volvo's car stoppage system went when at its launch when the car drove straight into the barrier.
Wasn't it a Benz?
For years and years the general motivation behind car development seems to be "let the car do it", rather than "teach the driver how to do it". It is such an awful mentality and is the first step to cars being physically driven being phased out and then banned due to being "un-safe".
Cars are perrsonal transportation vehicles, why would you need to control such a vehicle as your entire reason for being in one is to get where you want to go without anyone else in it (or a limited number of people that you like)? The enthusiast market is very small by comparison to the above so it is not a priority.
 
Cars are perrsonal transportation vehicles, why would you need to control such a vehicle as your entire reason for being in one is to get where you want to go without anyone else in it (or a limited number of people that you like)? The enthusiast market is very small by comparison to the above so it is not a priority.

Because without full control you lose the fact that cars essentially are a form of freedom. You can go anywhere there is a road, at a variety of speeds in full control of something you have bought. How long before cars essentially become driver-less taxis?
From that point whoever controls this system of transport has complete control of freedom of movement. I just plain don't like the idea of this in any way shape or form.

If they took the money invested in this and made public transport better andcheaper and driving tests harder and more involved it would solve the problem they say they are trying to solve. Of people who don't like, or want, to drive and as thus don't put the effort into learning how properly.
 
Top