Autonomous Vehicle (AV) Taxis - Yea, Nay or Maybe? - An ARRSEr Poll

Autonomous Vehicle (AV) Taxis - Yea, Nay or Maybe? - An ARRSEr Poll

  • Hell yes! Where do I sign?

    Votes: 17 30.9%
  • Umm…maybe let others be the guinea pigs first, will take a rain check.

    Votes: 14 25.5%
  • Hell no! This is the start of Skynet. Now, where is my Model T taxi?

    Votes: 24 43.6%

  • Total voters
    55

Cold_Collation

LE
Book Reviewer
#61
Maybe but to listen to the manufacturers and politicians who have fallen for the salesman's patter, it's only a single figure number of years away. T'ain't so, t'ain't so by a wide margin.

The city I live in is looking to install a Clean Air Zone and want £40 million from HMG to do so. That's just for some cameras, signs, road markings and the back office system to run the thing. Personally I don't know where they've got that figure from but it gives an indication of the scale of likely costs for implementing new technology.
Google said three years over a year ago. The guy behind Navteq, the digital mapping company bought up by Nokia and now part of HERE, reckons closer to 30 years. I'm erring towards the latter.
 
#62
Exactly. Throw a dark rainy night or an inch or two of slush/snow on the road surface and a bit of dense fog into the mix just for good measure, making the lane lines disappear and the edge of the carriageway indeterminable at best and it's game over time.

As a driver, I can get that half second glance of a child heading towards the road at speed at is then disappears behind a parked van, knowing that it's likely to run out in front of me from behind the parked van and start to brake/swerve in anticipation. The current level of tech can't do that.

Automatic parking in a tight spot - yes. Anything else - no, for the foreseeable future.

This might - stress might - work in US cities with their generally regular block layout but in the UK, forget it.
Yes it can, they will drive at a speed at which they can stop in time if a child pops out from behind a van.

Teslas can already be autonomous, the only thing holding it back is infastructure and legislation.
 

skid2

LE
Book Reviewer
#63
Great, just the thing for hitting the office and the off to a ‘meeting’, afterwards. Especially when the office is in town, no problem with parking. That’s got to be good.
I could see the return of the staff meeting

Edit but won’t we need our own special Britain navigation system with no US or Eu involvement.
 
#64
Teslas can already be autonomous, the only thing holding it back is infastructure and legislation.
A little more complicated than that. The software needs improving to a state where the vehicle can, safely, be allowed to drive itself. I believe that Tesla and other manufacturers state that the driver must always be 'in charge', hands on the steering wheel and feet on the pedals.
From testing I was involved in with NASA about 30 years ago it is extremely difficult to get a human to pay attention and concentrate on a task that they're only peripherally engaged in. When they do have to 'switch on' it takes time to orientate yourself and pick up on whats around you and whats going on before you can take action.

A Tesla spokeperson said: “When using autopilot, drivers are continuously reminded of their responsibility to keep their hands on the wheel and maintain control of the vehicle at all times.
“Tesla has always been clear that autopilot doesn’t make the car impervious to all accidents, and before a driver can use autopilot, they must accept a dialogue box which states that ‘autopilot is designed for use on highways that have a center divider and clear lane markings’.”
Failing to detect stopped vehicles is a known problem for makers of cars with autonomous features. Tesla’s manual warns: “Traffic-Aware Cruise Control cannot detect all objects and may not brake/decelerate for stationary vehicles or objects, especially in situations when you are driving over 50mph (80kmh) and in situations where a vehicle you are following moves out of your driving path and a stationary vehicle or object is in front of you.”
 
#65
A little more complicated than that. The software needs improving to a state where the vehicle can, safely, be allowed to drive itself. I believe that Tesla and other manufacturers state that the driver must always be 'in charge', hands on the steering wheel and feet on the pedals.
From testing I was involved in with NASA about 30 years ago it is extremely difficult to get a human to pay attention and concentrate on a task that they're only peripherally engaged in. When they do have to 'switch on' it takes time to orientate yourself and pick up on whats around you and whats going on before you can take action.

A Tesla spokeperson said: “When using autopilot, drivers are continuously reminded of their responsibility to keep their hands on the wheel and maintain control of the vehicle at all times.
“Tesla has always been clear that autopilot doesn’t make the car impervious to all accidents, and before a driver can use autopilot, they must accept a dialogue box which states that ‘autopilot is designed for use on highways that have a center divider and clear lane markings’.”
Failing to detect stopped vehicles is a known problem for makers of cars with autonomous features. Tesla’s manual warns: “Traffic-Aware Cruise Control cannot detect all objects and may not brake/decelerate for stationary vehicles or objects, especially in situations when you are driving over 50mph (80kmh) and in situations where a vehicle you are following moves out of your driving path and a stationary vehicle or object is in front of you.”
So pretty much infrastructure and legislation then?

I take your points though and they are valid but consider that the current live version of the Tesla OS does not represent the apex of their development. They are several version ahead in development but cannot release the latest version that deal with autonomous driving without first testing the hell out of it in a realistic environment and convincing the government that its safe.

We are almost there.
 
#66
So pretty much infrastructure and legislation then?
Did you miss this bit?
"Traffic-Aware Cruise Control cannot detect all objects and may not brake/decelerate for stationary vehicles or objects, especially in situations when you are driving over 50mph (80kmh) and in situations where a vehicle you are following moves out of your driving path and a stationary vehicle or object is in front of you.”
A long long way off if the thing cannot detect stationary objects right in front of it.
 
#68
Ah screw it. Looks like you crusty folk just want to go back to the old days.
No we just require to have something that's proven to be safe on the roads for all road users under all the conditions we currently drive in. I don't think that's too much of an ask.
 
#69
I't interesting that the accidents in which AVs have been involved, principally in the US, have mostly been caused by human error - the AVs have been rear-ended while trying to merge into moving traffic. Better human anticipation, or poor and impatient human driving? I err towards the latter.
Most? Pretty much all!

I did some research into AV's earlier in the year for implementing AV's into a warehouse environment. As there was no guidance or industry standards I had to think outside the box, so I looked at the US trials. In one state (California I think) All incidents had to be logged on a form ,and the form published to the net. My biggest gripe was the most common incident that occurred was when the human driver took control from the AI as a situation developed (invariably too late to effect the outcome), and so had to be discounted.

The sample wasn't huge (just under 50 incidents) but it did give a clear result. Some 67% of incidents happened when a 2nd vehicle rear ended the AV at traffic lights. The next biggest group was 21%, in it a 2nd vehicle was overtaking and fucked up causing a collision.

There were another two categories in it, but in all cases, bar one, the reason for the incident was human error on the driver of Vehicle 2.

That final one was a weird one though. An AV stopped at a pedestrian crossing. A male on the opposite side of the road then started shouting and gesticulating, he then charged at the stationary AV, running into it at full speed.
 
#70
Is the aim of this automation to make life

1, safer for the passenger
2, cheaper for the operator
3, an equally quick, safe and affordable journey for the passenger
4, a playground for tech developers using human guinea pigs

My bet is on 2, reliant on vast government ( ie tax payer) funded infrastructure.
 
#71
Did you miss this bit?
"Traffic-Aware Cruise Control cannot detect all objects and may not brake/decelerate for stationary vehicles or objects, especially in situations when you are driving over 50mph (80kmh) and in situations where a vehicle you are following moves out of your driving path and a stationary vehicle or object is in front of you.”
A long long way off if the thing cannot detect stationary objects right in front of it.
yes, they cant release the later version to the public before the go ahead is given.
 
#72
That final one was a weird one though. An AV stopped at a pedestrian crossing. A male on the opposite side of the road then started shouting and gesticulating, he then charged at the stationary AV, running into it at full speed.
Probably a crusty old ARRSEr here..;)
 
#73
There is and always will be, a big problem with AV and any AI, "proprietary software"! Until that problem is solved. , well, who knows? Want to put your life in the hands of one all dominating company?. Not that O2 is dominating but an example of what happens when tech fuckes up.
 
#74
There is and always will be, a big problem with AV and any AI, "proprietary software"! Until that problem is solved. , well, who knows? Want to put your life in the hands of one all dominating company?. Not that O2 is dominating but an example of what happens when tech fuckes up.

No O2 isn't but O2's failure this morning left me with limted communications and dysfunctional telematics for half the day
Not the end of the world but clearly demonstates that technology has its limitations. The same limitations and defects apply to all automotive electronics. Autonomous vehicles will break just the same as conventional ones do but they don't have the driver to deal with it.
 
#75
yes, they cant release the later version to the public before the go ahead is given.
And they can't get the go ahead until they prove its safe, a catch 22 ;-) There has to be a very large issue here in testing under realistic conditions with all of the idiot non AV drivers on the road. Also the realistic situation which TESLA are trying to guard against where you're following a car which does an avoid manoeuvre right smart and your AV doesn't. As per this video:
Tesla Model S on autopilot crashes into van on highway | Daily Mail Online
 
#76
There was a discussion about this some time ago.It seems the biggest obstacle to all this new tech in cars is the "insurance industry".
Deciding where to draw the line between human driver/automatic driver is still bogged down in red tape.
 

mercurydancer

LE
Book Reviewer
#77
My car has anti-collision in its bag of tricks. Very useful indeed, to the point of it being an essential bit of kit. it does self-parking too, but I dont use it much. This level of automation is certainly not autonomous driving, but it does help. I suspect that given a few years, the autonomous car will be a feature in some city centres. I look forward to it.
 
#78
And they can't get the go ahead until they prove its safe, a catch 22 ;-) There has to be a very large issue here in testing under realistic conditions with all of the idiot non AV drivers on the road. Also the realistic situation which TESLA are trying to guard against where you're following a car which does an avoid manoeuvre right smart and your AV doesn't. As per this video:
Tesla Model S on autopilot crashes into van on highway | Daily Mail Online
It's incredibly common at roadworks on a motorway/dual carriageway to have a cone taper in the nearside lane for short term works. It's also incredibly common to have a platoon of HGVs travelling in the nearside lane. The first one sees the taper ahead and pulls into the next lane. The second one reacts a little later and closer to the taper. As does the third. The forth one hits it, scattering cones, battery lamps and signs all over the place. I've no idea how an AV would react appropriately to that scenario.

"What about the warning signs on the approach to the closure?", I hear you say. You'd be amazed how many drivers don't see them and only change lane when they see the taper ahead of them. If they've got time to do so.

I believe in Holland they occasionally use temporary rumble strips laid in the to be closed lane 150m or so in advance of one of these things, the "Pick and place":

Pick & Place TMA Arrow Warning Unit (PPT)

If your AV doesn't recognise the rumble strips for what they are (because it might think they are just debris) and there's a couple of HGVs in front of you...
 
#79
Like it or not the creep began years ago to AV's, the first recognised safety critical item to partially take over was ABS! We now have adaptive cruise control, lane assist, fly by wire throttle and brakes. Auto gearboxes where the stick is just an electronic switch. Auto handbrakes and hill start assist. As mentioned above self-parking is now common, so the motors are in place for taking over steering. Collision control now steps in if the car thinks we're not paying attention. We are being engineered out of control of our vehicles slowly but surely.
 
#80
Yes it can, they will drive at a speed at which they can stop in time if a child pops out from behind a van.

Teslas can already be autonomous, the only thing holding it back is infastructure and legislation.
So in a 30, 40, 50 or 60 limit, how does it scan the road ahead, identify with at least 99% confidence the potential "hiding places" the kid could pop out from and adjust it's speed appropriately, without constantly speeding up and slowing down, bearing in mind the various stopping distances for those speeds? Unless it's going to travel at, say, 10mph everywhere, which will win it no friends amongst people who need to be places, particularly those stuck in the queue behind it.

It might work on a sunny day on a wide, straight US street. A typical UK street in January is an entirely different beast.

Oh, and you missed out "unaffordably expensive" in front of "infrastructure", (which you also spelt wrong :twisted:).

I'm a traffic engineer who has heard all the patter from salesmen trying to push the latest technology into my industry. All too often the latest thing is found to not live up to the salesman's promises.
 

New Posts

Latest Threads

Top