Oxford University's driver-less car

Use this board for general non-cycling-related chat, or to introduce yourself to the forum.
kwackers
Posts: 13748
Joined: 4 Jun 2008, 9:29pm
Location: Warrington

Re: Oxford University's driver-less car

Postby kwackers » 4 Apr 2012, 9:05am

JohnW wrote:
gaz wrote:
kwackers wrote:....The obvious fact is .................................that's what would really get peoples goat. :wink:


Just think how great this utopian future would be if we also get the riderless bicycle and pedestrianless shoes. :wink:


Nice one Gaz.

Not quite. A riderless bicycle transports nothing, on the other hand you could make something like a Segway with automatic navigation. :wink:

Cyclists though either ride bikes to be green/save money/for exercise or for fun. Doesn't really matter what technologies exist these can't be replaced.

kwackers
Posts: 13748
Joined: 4 Jun 2008, 9:29pm
Location: Warrington

Re: Oxford University's driver-less car

Postby kwackers » 4 Apr 2012, 9:10am

JohnW wrote:and my belief is that it gives me the common sense to believe that a driverless car shouldn't ever happen. Couldn't ever happen? - well, a hundred years ago no-one would credit that man would actually set foot on the moon - or have a single weapon that could destroy all life - but the accademics and the scientists have created one. Couldn't ever happen? - I'm sure it could, but hang your bike up when you see it coming!

Just to take you up on this point. The danger to cyclists isn't driverless cars, if and when they happen they'll be much better than your average driver at dealing with us.
The danger is the intermediate stage, in fact the one that happening now where some of the technology is filtering down in an immature state (i.e. collision avoidance) whilst still being primarily under the control of a human.
There is the real danger that people will rely on the technology to 'help' them out and thus pay less attention to the roads.

If you want to hang up your bicycle I'd suggest now is a good time, get it back out in a few years when the technology is advanced enough to be reliable and the decision to avoid something is taken completely out of human control.

User avatar
meic
Posts: 19355
Joined: 1 Feb 2007, 9:37pm
Location: Caerfyrddin (Carmarthen)

Re: Oxford University's driver-less car

Postby meic » 4 Apr 2012, 9:20am

If you go back and look at the "Tomorrow's World" programmes of the 1970s it becomes clear that the future never develops as planned. The increase in computer power and harnessing it has been the biggest unpredicted change.
So instead of super efficient vehicles human controlled on linear rails we are likely to end up with inefficient gas guzzlers with computer brains transporting morons.

I think that Kwackers is forgetting that humans and their politicians will manage to make an utter mess of his utopic vision. Then good old "Garbage In, Garbage Out" will also cause a lot of hiccups for the system.

Is there a good reason why the Rail Authorities have not taken this on board?
Yma o Hyd

kwackers
Posts: 13748
Joined: 4 Jun 2008, 9:29pm
Location: Warrington

Re: Oxford University's driver-less car

Postby kwackers » 4 Apr 2012, 9:41am

meic wrote:I think that Kwackers is forgetting that humans and their politicians will manage to make an utter mess of his utopic vision. Then good old "Garbage In, Garbage Out" will also cause a lot of hiccups for the system.

GIGO is a dated term and less applicable these days particularly to the sort of systems we're talking about here.
When applied to sensory systems humans suffer exactly the same issues (take a look on the hi-vis thread for a start) where computer sensory systems improve on this is that they don't rely on a single system but multiple, with backup and with proper fail safe (no more driving when blinded by the sun ;-) ).


meic wrote:Is there a good reason why the Rail Authorities have not taken this on board?

What, like the London Underground?
I think you'll find there are plenty of driverless train systems around, the only real issues they suffer from are usually union related...

thirdcrank
Posts: 28648
Joined: 9 Jan 2007, 2:44pm

Re: Oxford University's driver-less car

Postby thirdcrank » 4 Apr 2012, 9:42am

meic

While your Tomorrow's World observation is spot on, I think you are applying this the wrong way round in apparently assuming that the present trends must continue.

Apart from that, it seems to me that most of the apparent shortfalls of this technology stem from assuming it would have to cope with the present environment and couldn't, but that's not how things do develop. It seems silly to point this out, but barges couldn't run on roads and canals were cut. Trains couldn't run on either so tracks were laid. It seems to me that the impreatives of more efficient use of fuel will eventually compel significant changes in travel modes.

kwackers
Posts: 13748
Joined: 4 Jun 2008, 9:29pm
Location: Warrington

Re: Oxford University's driver-less car

Postby kwackers » 4 Apr 2012, 9:50am

Incidentally my utopian view of the future doesn't involve personal vehicles - computer controlled or not.

Where I'd hope this technology is most likely to find use is as a personal vehicle replacement.
I'd envisage better mass transit systems with a much smaller fleet of 'personal' vehicles which are shared. A bit like Boris's bikes in London but electric vehicles for short (say less than 10 miles) journeys.
Instead of private cars you simply call one from your smart phone and it rolls up from nearby storage, you tell it where you want to go and off it sets...

Such a system would be cheap and personal, an ideal replacement for private vehicles.
If I were a producer for tomorrows world that's the idea I'd pitch. Will it happen? Probably not but I live in hope...

User avatar
meic
Posts: 19355
Joined: 1 Feb 2007, 9:37pm
Location: Caerfyrddin (Carmarthen)

Re: Oxford University's driver-less car

Postby meic » 4 Apr 2012, 10:14am

I try not to get caught in terms of present or even local (National or Western to the more inward looking people) trends and keep an eye on the permanent and universal trends.

Radio shows like "In our times" show how the same old themes reappear throughout history of the world. The future will probably be very much like the present for most of the population and the differences will be due to unforeseeable developments.

My "garbage in garbage out" reference was in response to the suggestion that the computerised vehicles would be informed about traffic and weather conditions.

Though I can see that computer units that were all communicating their intentions to each other would move a lot more efficiently than cars whose drivers cant be bothered to indicate at junctions.
Yma o Hyd

User avatar
RickH
Posts: 4540
Joined: 5 Mar 2012, 6:39pm
Location: Horwich, Lancs.

Re: Oxford University's driver-less car

Postby RickH » 4 Apr 2012, 10:21am

kwackers wrote:Incidentally my utopian view of the future doesn't involve personal vehicles - computer controlled or not.

Where I'd hope this technology is most likely to find use is as a personal vehicle replacement.
I'd envisage better mass transit systems with a much smaller fleet of 'personal' vehicles which are shared. A bit like Boris's bikes in London but electric vehicles for short (say less than 10 miles) journeys.
Instead of private cars you simply call one from your smart phone and it rolls up from nearby storage, you tell it where you want to go and off it sets...

Such a system would be cheap and personal, an ideal replacement for private vehicles.
If I were a producer for tomorrows world that's the idea I'd pitch. Will it happen? Probably not but I live in hope...

Do you mean something along the lines of the ones in operation around Heathrow (also ITN video on Youtube)? Harder to scale up to a mixed environment shared with non-automated transport I know but combined with something like the Bosch collision avoidance system it could be feasible. Out "in the wild" they would probably have to (at least initially) operate like trams on fixed routes, which could be either segregated or shared.

Rick.

JohnW
Posts: 6236
Joined: 6 Jan 2007, 9:12pm
Location: Yorkshire

Re: Oxford University's driver-less car

Postby JohnW » 4 Apr 2012, 12:01pm

kwackers wrote:
JohnW wrote:If driverless cars were allowed on the roads, what would be the point? To eliminate drivers? If not to eliminate drivers, then why?

..............1> Safety, driverless cars don't break rules, respond faster to 'situations' and are more aware of their environment.........................A driverless car can be an office on the way to wherever you're going. No problems taking phone calls, using computers or even watching a film..........................A driverless car will move people from A to B regardless of their ability to drive. From small children to going and getting your 100 year old grandma...........................


Well Kwackers - who's going to 'supervise' it? The impression that I had from one of your previous posts was that, even if a driverless car did have weaknesses, all would be well because it would be "supervised". Supervised by whom? - and how then? - if they're watching a film.

Now c'mon kwackers - there are plenty of drivers who get into unbelievable schtuck just by trusting satnavs and driving as they instruct. I submit that satnavs are not an acceptable substitute for knowing your way.

As for being "more aware of their environment" - you'll have to make one heck of an argument to sustain that proposition - environment is a human thing. You seem to believe that I could be the "...............sort of person who thinks the point of a car is to drive it............." - no I don't - I've never had a car, learned to drive 45 years ago but didn't take a test and never driven since. I don't wish to do any harm.

How do you know that driverless cars don't break rules? - how many of them are there? - what are the rules? - who will make the rules?.......and so it goes on. If someone is "developing" a driverless car, then there's neither experience nor data for you to base you proclamations upon

I read somewhere once - it must have been 30-35 years ago now, so numbers may have changed - that 80% of research worldwide was devoted to armaments and weaponry - i.e. killing people - rather than improving quality of life.

No change there then.

There's no solidarity or credibility in your arguments in this matter Kwackers, and I think that you're arguing for the sake of it.

Driverless cars may very well come - scientists can't rest without doing harm - but hopefully not in my day.

(For the avoidance of doubt, I use the term "arguing" more in the sense of debating, and not - repeat not - in the sense of quarrelling).

This is my third edit of this post. This one and the previous two were made to correct spelling mistakes. I'm only human......
Last edited by JohnW on 4 Apr 2012, 12:28pm, edited 3 times in total.

gilesjuk
Posts: 3270
Joined: 17 Mar 2008, 10:10pm

Re: Oxford University's driver-less car

Postby gilesjuk » 4 Apr 2012, 12:02pm

kwackers wrote:Utter nonsense.
Do you fly? If so you spend most of your time at the mercy of computers.


Fly by wire is incredibly simple. There's a damn good reason for that too, planes fly through the air with no obstacles other than other planes which are kept far away in the flight plan.

All a fly by wire system needs to do is measure the distance it is from the flight plan and move back to the flight plan as it drifts slightly away. This is like when driving a car around a corner, you don't look at the corner and then move the wheel to the exact position required and then back again, you make a series of small adjustments as required during the move.

I work in the world of software and I know that I wouldn't ever put my life in the hands of complex software when it comes to driving. There are just too many things that can go wrong, hardware faults, software faults and oversights. A bit of bright sunlight can make some sensors useless. ECUs already fail and have problems, such as the Nissan Micra ECU that wasn't switched off when the ignition was switched off, flattening the battery. I just don't trust motor manufacturers to keep me safe.

Technology assistance often makes people worse at any given task. It becomes a crutch.

thirdcrank
Posts: 28648
Joined: 9 Jan 2007, 2:44pm

Re: Oxford University's driver-less car

Postby thirdcrank » 4 Apr 2012, 12:06pm

I wasa going to mention commercial flying in my earlier post and I was distracted. It's easy to dismiss this as an example because of the very obvious need for expert flight crew, but I think that misses the point. On the ground, the appropriate action in any sort of emergency / unexpected situation is to stop, then sort it out. This is a luxury not available to the people in an aeroplane.

gilesjuk
Posts: 3270
Joined: 17 Mar 2008, 10:10pm

Re: Oxford University's driver-less car

Postby gilesjuk » 4 Apr 2012, 12:06pm

[XAP]Bob wrote:The google cars have done many many miles in california - no accidents whilst driven by computer, which is more than can be said for the fleshy option.


They have jaywalking laws in the US, so no people running across the road to worry about.

I think letting the computer take over on motorways may work. But for cities and towns there's too much going on.
Last edited by gilesjuk on 4 Apr 2012, 12:07pm, edited 1 time in total.

JohnW
Posts: 6236
Joined: 6 Jan 2007, 9:12pm
Location: Yorkshire

Re: Oxford University's driver-less car

Postby JohnW » 4 Apr 2012, 12:07pm

gilesjuk wrote:
kwackers wrote:Utter nonsense.
Do you fly? If so you spend most of your time at the mercy of computers.


Fly by wire is incredibly simple. There's a damn good reason for that too, planes fly through the air with no obstacles other than other planes which are kept far away in the flight plan.

All a fly by wire system needs to do is measure the distance it is from the flight plan and move back to the flight plan as it drifts slightly away. This is like when driving a car around a corner, you don't look at the corner and then move the wheel to the exact position required and then back again, you make a series of small adjustments as required during the move.

I work in the world of software and I know that I wouldn't ever put my life in the hands of complex software when it comes to driving. There are just too many things that can go wrong, hardware faults, software faults and oversights. A bit of bright sunlight can make some sensors useless. ECUs already fail and have problems, such as the Nissan Micra ECU that wasn't switched off when the ignition was switched off, flattening the battery. I just don't trust motor manufacturers to keep me safe.

Technology assistance often makes people worse at any given task. It becomes a crutch.


Nice one Giles.

kwackers
Posts: 13748
Joined: 4 Jun 2008, 9:29pm
Location: Warrington

Re: Oxford University's driver-less car

Postby kwackers » 4 Apr 2012, 12:43pm

gilesjuk wrote:I work in the world of software and I know that I wouldn't ever put my life in the hands of complex software when it comes to driving.

Me too. I've written many AI simulations to move objects through complex worlds. It's an iterative problem and fairly simple to solve. I'd have absolutely no problem allowing a tried and tested system on the road.

There are just too many things that can go wrong, hardware faults, software faults and oversights.

Hence redundancy and self checking systems. It's also a nonsense to claim you cant factor these in yet ignore the *many* faults that humans have, from simply dying at the wheel foot hard on the accelerator to ignoring traffic conditions, signals or simply deciding to drive like a fool for a while because "they felt like it".

A bit of bright sunlight can make some sensors useless. ECUs already fail and have problems, such as the Nissan Micra ECU that wasn't switched off when the ignition was switched off, flattening the battery. I just don't trust motor manufacturers to keep me safe.

Indeed bright sunlight makes eyes useless (as evidenced by the many threads we get on cyclists killed by such a factor). The difference is a computer will use other sensors or if all fails simply shut down to a fail safe mode (i.e. stop) rather than allow their ego/stupidity to allow them to continue.

Technology assistance often makes people worse at any given task. It becomes a crutch.

Only when it's an assistance rather than a substitute. As I said above the danger is now when immature "assist" systems are added to existing vehicles.

The nice things about computers and automated systems is that if there is a problem it can be fixed, not only that but long before the system became wholesale the problems would have been found and fixed - hence the OP and the fact the google car has been driving around for a while.
Can you say the same for humans? A creature designed with a fairly sedate top speed and 80kg mass suddenly expected to be able to respond properly at high speed and with significant mass? Until genetic engineering of humans becomes possible then we're stuck with what we've got along with the significant road casualties currently inflicted precisely because we're not only fallible but extremely so.

User avatar
[XAP]Bob
Posts: 16941
Joined: 26 Sep 2008, 4:12pm

Re: Oxford University's driver-less car

Postby [XAP]Bob » 4 Apr 2012, 1:27pm

gilesjuk wrote:
[XAP]Bob wrote:The google cars have done many many miles in california - no accidents whilst driven by computer, which is more than can be said for the fleshy option.


They have jaywalking laws in the US, so no people running across the road to worry about.

I think letting the computer take over on motorways may work. But for cities and towns there's too much going on.


Two points - fleshbags don't always obey laws, animals never knowingly do so.

Point three - California doesn't have a "cross anywhere = jaywalking" law.
A shortcut has to be a challenge, otherwise it would just be the way. No situation is so dire that panic cannot make it worse.
There are two kinds of people in this world: those can extrapolate from incomplete data.