Connected and self-driving cars are being sent to Coventry The UK Autodrive connected car consortium will start practical trials of its connected and autonomous cars on the streets of Coventry, it has declared. Jaguar Land Rover (JLR), Ford and Tata Motors European Technical Centre (TMETC), all of which are partners in the project, are trialling a number of connected car features in …
The ring road itself is fine (apart possibly for junction 2) if people take notice of the signs and use it how it is designed.
As for all the silly shared space junctions inside the ring road that is a different matter.
They are also in the process of 'upgrading' a load of the traffic lights at busy junctions outside the ring road, though if anything their 'smart system' appears to have made the congestion worse at the one I pass through every day.
...as any other place to smash and crash cars not ready for prime time. There are NO mandatory hardened security, certified maintenance, agreed liability or other necessary aspects of "connected cars" including EVs. When the software is written to try and avoid an unavoidable accident, the programmer has to decide who will die in an unavoidable accident and there are no legal agreements. Without redundant and failsafe systems many people will die. Federal governments are taking the position of reacting after people are injured or killed as has already happen with drivers using Tesla's "Autopilot". The Feds INMHO are abdicating their responsibility to insure public safety by allowing rush-to-market vehicles so that the first suppliers of these vehicles can reap massive profits. It's simply unacceptable and complete irresponsible to not mandate the appropriate vehicle safety, security, design and maintenance requirements before these vehicles are allowed to be tested on roadways.
Liability has been sorted on this side of the pond actually. Driverless cars were heavily debated in parliament this year.
Also, there's no such thing as an unavoidable accident between two cars. It's that kind of thinking that makes me want these sooner rather than later so we can get human drivers off the road. Humans always ignore the third option when considering whether the driver or pedestrian should die. Option 3 = drive slower! Machines don't consider "late for work" in their calculations...
"Also, there's no such thing as an unavoidable accident between two cars. "
So a high speed blow-out never occurs?
What about hitting one of the mother-of-a-pothole that are appearing with increasing frequency and ripping a wheel off?
What about general mechanical failure?
There *should* be fewer accidents as many driving accidents are avoidable however there's no such thing as unavoidable unless you never take the vehicle out of the drive (even then you can get killed by handbrake failure ... but I suppose that's avoidable by not parking it on a slope?)
"the programmer has to decide who will die in an unavoidable accident "
I don't know where this strange notion comes from. If an accident is unavoidable the primary concern of the software should be the safety of the passengers. There is no 'correct' solution to the trolley problem, everyone has their own internal answer, which likely in the real world (ie if it actually happens with them behind the wheel) is different from their 'declared' answer from the safety of one's sofa. Why should we expect that AIs should satisfactorily solve the trolley problem, or programmers do so on their behalf?
If, as would make the sense, autonomous vehicles make extensive use of "sat nav" and mapping data, the ridiculous errors everyone likes to keep pointing should be no issue. There's absolutely no reason why a the maps can't rapidly update to exclude the tiny single track lanes, unclassified roads, and low bridges as those errors are encountered. Just as with live traffic data, road type and suitability for the vehicle in question can be rapidly learnt by what will undoubtedly be largely cloud based systems updated by millions of drones (us).
Most of the existing issues are caused by locally stored very outdated maps on primitive devices that provide little or no options for feedback. And in some case, idiots... I would expect a "smart" car to be able to spot a cliff or a river...
The problem here is that UK road design and management has for *years* taken a stance that private motoring is bad, and not to be encouraged.
Hence the proliferation of road layouts and traffic signals which actually *slow* a journey down.
If you want to win a bet, compare the A38 from Longbridge to Edgbaston priory via the "bypasses" which takes about 3-4 minutes longer (watch out for the *pedestrian lights with their 20-second phase) than going through the town centres.
Ever sat waiting for a light to change, and noticed that the pedestrian lights 25m on seem to be such that you hit them at red as soon as your light is green ? You didn't really think that's an accident.
...but why is a trial for driverless cars being announced by the Chancellor in a pre-budget leak? What has it to do with tax on beer, national insurance and government department spending?
Apparently, "up to £1 billion" will be available for hi-tech projects "including £75m for research on artificial intelligence, £400m for electric car charging points and £100m to boost clean car sales" according to the Guardian. So he is not really spending anything on AVs.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
