Autonomous cars? I think not

User avatar
Ross
Posts: 5380
Joined: Sat Nov 07, 2009 8:53 pm

Re: Autonomous cars? I think not

Postby Ross » Sat Aug 11, 2018 7:57 am

mikesbytes wrote:Of course the Australian version of the Autonomous cars will need to customised to match current driving expectations - settings;
1. MGIF
2. Scream up wrong lane and push in near the front
3. Accelerate on orange lights that don't have speed/red light cameras
4. Blast horn if vehicle in front doesn't commence within [user selectable] 10'ths seconds from light turning green
5. Park on no stopping if closer to coffee shop
6. Ignore one metre rule
7. Ignore advisory signs
8. Wind windows down and play doof doof music full blast
Retiree options;
9. Get confused between the accelerator and brake pedals
10. Drive at 50kph in an 60kph zone
11. Drive at 50kph in an 80kph zone
12. Drive at 50kph in an 100kph zone
13. Park 800mm from the kerb


Gold! :mrgreen:

User avatar
Strawburger
Posts: 1710
Joined: Tue Aug 26, 2008 12:19 pm
Location: Dulwich Hill, Sydney

Re: Autonomous cars? I think not

Postby Strawburger » Sat Aug 11, 2018 5:53 pm

Ross wrote:
Strawburger wrote:
Ross wrote:Not sure how autonomous cars will react around roadworks where there are just plastic cones set out to mark the "road" and a roadworker with a "lollipop" stop/go sign and a reduced speed limit? The traffic cones are often just lazily set out and not really in a tight or symetrical configuration.

Often there are signs out a km or so ahead of roadworks warning/alerting motorists of roadworks ahead, how will autonomous cars deal with this as they aren't able to read (it could be argued no differently to motorists who ignore/don't see them because of mobile phone use/other distraction...).

Will each traffic ccone require an embeded chip so te austonomous vehicle recognises what it is? Will the temporary speed limit signs also need an ebedded chip so the car knows to drive at 40km/h instead of 100km/h? What about the stop/go sign?


Additional measures to ensure the vehicle will be able to read what it needs to. It would just be a modification to the traffic management process


That reply reads like a govt press release. Tells us nothing. I'm interested in a real world actual processes answer.


Sensors/ iot devices installed into traffic management signage or cones/barriers to warn the vehicles that an alternate layout is in operation.

In the process of applying for a road occupancy licence, the traffic management plans would require an additional submission to the various digital mapping authorities with the changes.

Lastly, the manual control would not be in effect as there are already solutions to remove the humans from lollipop type traffic controls. Most of the big traffic control companies have this on the major highways and this will cascade down by the time autonomous vehicles are on the open roads.

Any more details you would like to know? I’ll try to answer to the best of my knowledge.

Keep in mind that technology is rapidly evolving and technology mentioned now and will be replaced by stuff that doesn’t exist right now
n=10 (2013 & 2004 roads,2010 track,2x 2009 foldups,1990 hybrid,1992 trainer,2007 rental,1970's step through,1980's zeus)

User avatar
antigee
Posts: 552
Joined: Sun Sep 01, 2013 10:58 am
Location: just off the Yarra Trail but not lurking in the bushes

Re: Autonomous cars? I think not

Postby antigee » Sat Aug 11, 2018 10:33 pm

the issue is that a lot of current generation automonous vehicles aren't - they are detecting things like other vehicles within a defined already mapped route

cones with sensors - just like cyclists with sensors - cheap and easy solution to a new business model but somewhere in there is a problem and the solution is more restrictions for cyclists (and pedestrians).....

https://www.bikebiz.com/landscape/bike-beacons

"...............Many lives would be saved if cyclists used “bicycle-to-vehicle” sensors, a cycle industry association boss told automotive leaders at an event on the first day of the Geneva Motor Show.
Manuel Marsilio, general manager of the Confederation for the European Bicycle Industry was speaking at the ‘Future Networked Car’ symposium................

It is the goal of the “connected car” industry to make cyclists use sensors or beacons so they can be detected more easily........."

User avatar
Howzat
Posts: 803
Joined: Wed Aug 15, 2012 7:08 pm

Re: Autonomous cars? I think not

Postby Howzat » Mon Aug 13, 2018 10:16 am

antigee wrote:It is the goal of the “connected car” industry to make cyclists use sensors or beacons so they can be detected more easily........."[/i]


Before: "Autonomous cars are here - one will drop the kids at school while you read a book on the way to work! What will you do with the all the extra time?"

Now: "Autonomous cars will work great and only require retrofitting the entire road network and everything on it with compatible technology to help them avoid running into things!"

User avatar
antigee
Posts: 552
Joined: Sun Sep 01, 2013 10:58 am
Location: just off the Yarra Trail but not lurking in the bushes

Re: Autonomous cars? I think not

Postby antigee » Tue Aug 14, 2018 1:00 pm

good news (?! ;-) ) CNBC report Uber are restarting trials soon albeit it in Pittsburgh not Tempe, Arizona

thats at the end of this piece about geofenced self driving loop mini buses in Frisco, Texas - the other good news is :

..... "(the) self-driving vehicle being tested as part of a new pilot program in Texas is designed not to blend in but to stand out — and in a big way. Its bright orange exterior is emblazoned with the words “Self-Driving Vehicle,” and it features billboard-like LED displays that alert nearby motorists and pedestrians with heads-up messages like “going now, don’t cross” and “waiting for you.”

so that sorts out a lot of my concerns

article: https://www.nbcnews.com/mach/science/wh ... ncna897151

User avatar
bychosis
Posts: 5706
Joined: Sat Jan 14, 2012 1:10 pm
Location: Lake Macquarie

Re: Autonomous cars? I think not

Postby bychosis » Tue Aug 14, 2018 2:51 pm

Strawburger wrote:
Ross wrote:
Strawburger wrote:
Additional measures to ensure the vehicle will be able to read what it needs to. It would just be a modification to the traffic management process


That reply reads like a govt press release. Tells us nothing. I'm interested in a real world actual processes answer.


Sensors/ iot devices installed into traffic management signage or cones/barriers to warn the vehicles that an alternate layout is in operation.

In the process of applying for a road occupancy licence, the traffic management plans would require an additional submission to the various digital mapping authorities with the changes.

Lastly, the manual control would not be in effect as there are already solutions to remove the humans from lollipop type traffic controls. Most of the big traffic control companies have this on the major highways and this will cascade down by the time autonomous vehicles are on the open roads.

Any more details you would like to know? I’ll try to answer to the best of my knowledge.

Keep in mind that technology is rapidly evolving and technology mentioned now and will be replaced by stuff that doesn’t exist right now

Having some experience with council and local roads maintenance, we have a very long way to go before all local road maintenance is managed in the same way they do for state roads. It is more likely that autonomous cars will be taught to read the road as we currently do. Sure, they'll be able to be supplemented by sensors and pre planned traffic management systems uploaded to the network, but the reality is that dogs and horses and flat batteries and those weird spots where your phone just wont work properly will mean that an autonomous car must rely on what it can see using lidar/radar/camera images and using the on board computer to interpret.

We are currently in a transition phase where more autonomous features are built into car and making us lazier as drivers. AEB, lane alert, cruise control that slows down when you follow a slower car are all features that serve only to reduce the need for a driver to pay attention, when driving that great big lump of steel should require full attention. This leads to additional distractions from the inevitable phone/device while we wait for a beep to occur to tell us to do something becuase the car is 'doing it for us'. Humans cannot remain alert to a situation when they don't need to (becuase the machine is). I think a driver in an old car with no fancy automatic features, no 'infotainment' system and the windows down is going to be more alert to the surrounding situation than a cocooned air conditioned automatic driver with all the latest gizmos.
Last edited by bychosis on Tue Aug 14, 2018 2:57 pm, edited 1 time in total.
bychosis (bahy-koh-sis): A mental disorder of delusions indicating impaired contact with a reality of no bicycles.

User avatar
bychosis
Posts: 5706
Joined: Sat Jan 14, 2012 1:10 pm
Location: Lake Macquarie

Re: Autonomous cars? I think not

Postby bychosis » Tue Aug 14, 2018 2:52 pm

antigee wrote:good news (?! ;-) ) CNBC report Uber are restarting trials soon albeit it in Pittsburgh not Tempe, Arizona

thats at the end of this piece about geofenced self driving loop mini buses in Frisco, Texas - the other good news is :

..... "(the) self-driving vehicle being tested as part of a new pilot program in Texas is designed not to blend in but to stand out — and in a big way. Its bright orange exterior is emblazoned with the words “Self-Driving Vehicle,” and it features billboard-like LED displays that alert nearby motorists and pedestrians with heads-up messages like “going now, don’t cross” and “waiting for you.”

so that sorts out a lot of my concerns

article: https://www.nbcnews.com/mach/science/wh ... ncna897151

Yay, more high-vis and visual pollution!
bychosis (bahy-koh-sis): A mental disorder of delusions indicating impaired contact with a reality of no bicycles.

User avatar
Thoglette
Posts: 4308
Joined: Thu Feb 19, 2009 1:01 pm

Re: Autonomous cars? I think not

Postby Thoglette » Tue Aug 14, 2018 3:14 pm

bychosis wrote: I think a driver in an old car with no fancy automatic features, no 'infotainment' system and the windows down is going to be more alert to the surrounding situation than a cocooned air conditioned automatic driver with all the latest gizmos.

That's not an "I think" :-)

The only time that starts to drift towards "contestable" is once the driver of the old vehicle starts to get fatigued due to extra NVH - after several hours on rutted roads.
Stop handing them the stick! - Dave Moulton
"People are worthy of respect, ideas are not." Peter Ellerton, UQ

User avatar
mikesbytes
Super Mod
Super Mod
Posts: 19239
Joined: Thu Dec 01, 2005 11:42 pm
Location: Tempe, Sydney
Contact:

Re: Autonomous cars? I think not

Postby mikesbytes » Tue Aug 14, 2018 10:56 pm

antigee wrote:"...............Many lives would be saved if cyclists used “bicycle-to-vehicle” sensors, a cycle industry association boss told automotive leaders at an event on the first day of the Geneva Motor Show.
Manuel Marsilio, general manager of the Confederation for the European Bicycle Industry was speaking at the ‘Future Networked Car’ symposium................

It is the goal of the “connected car” industry to make cyclists use sensors or beacons so they can be detected more easily........."


The cyclist was not wearing a sensor......
If the R-1 rule is broken, what happens to N+1?

User avatar
Strawburger
Posts: 1710
Joined: Tue Aug 26, 2008 12:19 pm
Location: Dulwich Hill, Sydney

Re: Autonomous cars? I think not

Postby Strawburger » Thu Aug 16, 2018 12:30 pm

Howzat wrote:
Before: "Autonomous cars are here - one will drop the kids at school while you read a book on the way to work! What will you do with the all the extra time?"

Now: "Autonomous cars will work great and only require retrofitting the entire road network and everything on it with compatible technology to help them avoid running into things!"


I would imagine that was a similar conversation had when the automobile started driving on the horse and cart / bicycle roads!
n=10 (2013 & 2004 roads,2010 track,2x 2009 foldups,1990 hybrid,1992 trainer,2007 rental,1970's step through,1980's zeus)

User avatar
Strawburger
Posts: 1710
Joined: Tue Aug 26, 2008 12:19 pm
Location: Dulwich Hill, Sydney

Re: Autonomous cars? I think not

Postby Strawburger » Thu Aug 16, 2018 12:35 pm

bychosis wrote:
We are currently in a transition phase where more autonomous features are built into car and making us lazier as drivers. AEB, lane alert, cruise control that slows down when you follow a slower car are all features that serve only to reduce the need for a driver to pay attention, when driving that great big lump of steel should require full attention. This leads to additional distractions from the inevitable phone/device while we wait for a beep to occur to tell us to do something becuase the car is 'doing it for us'. Humans cannot remain alert to a situation when they don't need to (becuase the machine is). I think a driver in an old car with no fancy automatic features, no 'infotainment' system and the windows down is going to be more alert to the surrounding situation than a cocooned air conditioned automatic driver with all the latest gizmos.


Totally agree. Humans are built to be lazy, it’s a struggle to keep the human engaged with the machine in semi autonomous mode now!
n=10 (2013 & 2004 roads,2010 track,2x 2009 foldups,1990 hybrid,1992 trainer,2007 rental,1970's step through,1980's zeus)

User avatar
Comedian
Posts: 5524
Joined: Mon Aug 09, 2010 7:35 pm
Location: Brisbane

Re: Autonomous cars? I think not

Postby Comedian » Sat Aug 18, 2018 8:01 am

Lukeyboy wrote:
mikesbytes wrote:
Tequestra wrote:Image

The driver was asleep at the wheel. If my 94 yo Grand Mother was driving that Volvo, that poor cyclist would have walked away.

http://www.abc.net.au/news/2018-03-22/s ... ed/9575624


Enough time/distance to swerve


Uber is going to be in big trouble as the car is at fault - technically speaking - because these cars use laser scanners. Lasers do not need light to detect objects. The sensor on a video camera needs light to be able to record what it sees. Lasers don't. If its lasers can fail to pick up the woman crossing the road then there can be massive loop holes in the software and or in its fundamental design.

This is exactly why I think autonomous cars are going to be 10 years off for far longer than we think.

If you crash your car - that's your fault. Nobody to blame but the drivers in question. If your autonomous car crashes - then the manufacturer is on the hook. What about if it's a multi vehicle crash - and the is a question of contribution from the autonomous car. The car makers aren't going to want that. So your driverless car did something wrong on the road, and charges need to be laid. Who gets charged? The CEO of the car company? I think we're going to need some kind of framework that absolves them of their legal responsibilities. That alone is going to take a long time to sort out.

I also think that the people who sell vehicles don't really want car sharing. So I think we're going to see increasingly complex driver aids - but I can't see the big players rushing to autonomy. They are happy with every family having four cars.

User avatar
mikesbytes
Super Mod
Super Mod
Posts: 19239
Joined: Thu Dec 01, 2005 11:42 pm
Location: Tempe, Sydney
Contact:

Re: Autonomous cars? I think not

Postby mikesbytes » Sat Aug 18, 2018 1:45 pm

If the R-1 rule is broken, what happens to N+1?

User avatar
Thoglette
Posts: 4308
Joined: Thu Feb 19, 2009 1:01 pm

Why driverless vehicles should not be given unchecked access to our cities

Postby Thoglette » Thu Sep 13, 2018 6:53 pm

The Conversation: Why driverless vehicles should not be given unchecked access to our cities
Authors:
Dominic Stead, Delft University of Technology
Anthony Kimpton, Casual Lecturer in Urban Sociology and Geography, The University of Queensland
Dorina Pojani, Senior Lecturer in Urban Planning, The University of Queensland
Iderlina Mateo-Babiano, Senior Lecturer in Urban Planning, University of Melbourne
Jonathan Corcoran, Professor, School of Earth and Environmental Sciences, The University of Queensland
Neil Sipe, Professor of Urban and Regional Planning, The University of Queensland

Autonomous, or driverless, vehicles can support and promote active travel, such as walking and cycling, when two basic conditions are met:

their access to cities is restricted
their use is pooled.
In the absence of these two conditions, autonomous vehicles could lead to a decline in active travel in cities and an increase in economic, social and environmental costs. Potential costs are rarely mentioned in the rhetoric about autonomous vehicles, much of which is highly optimistic.
Stop handing them the stick! - Dave Moulton
"People are worthy of respect, ideas are not." Peter Ellerton, UQ

User avatar
Alex Simmons/RST
Expert
Posts: 4642
Joined: Fri May 02, 2008 3:51 pm
Contact:

Re: Autonomous cars? I think not

Postby Alex Simmons/RST » Fri Sep 14, 2018 5:38 am

bychosis wrote:I think a driver in an old car with no fancy automatic features, no 'infotainment' system and the windows down is going to be more alert to the surrounding situation than a cocooned air conditioned automatic driver with all the latest gizmos.

I think we need to look at the data on accident/incident rates to make that assessment.

User avatar
bychosis
Posts: 5706
Joined: Sat Jan 14, 2012 1:10 pm
Location: Lake Macquarie

Re: Autonomous cars? I think not

Postby bychosis » Fri Sep 14, 2018 7:34 am

Alex Simmons/RST wrote:
bychosis wrote:I think a driver in an old car with no fancy automatic features, no 'infotainment' system and the windows down is going to be more alert to the surrounding situation than a cocooned air conditioned automatic driver with all the latest gizmos.

I think we need to look at the data on accident/incident rates to make that assessment.

Of course there are other significant factors like brakes, handling, tyres that reduce the safety of the vehicle. The behaviour of the driver will be affected by their risk perception, of which a large part comes from being alert to the environment.
bychosis (bahy-koh-sis): A mental disorder of delusions indicating impaired contact with a reality of no bicycles.

User avatar
redsonic
Posts: 1082
Joined: Wed Sep 26, 2012 8:08 pm
Location: Brisbane

Re: Autonomous cars? I think not

Postby redsonic » Fri Sep 21, 2018 11:52 am

Devonport authorities deliberately derailed an out of control driverless freight train in order to stop it. 2 pedestrians were injured:

ABC News

Shows a lack of failsafes/ foresight that they had no means to stop it safely.

Jmuzz
Posts: 567
Joined: Mon Jan 29, 2018 1:42 pm

Re: Autonomous cars? I think not

Postby Jmuzz » Fri Sep 21, 2018 12:17 pm

redsonic wrote:Shows a lack of failsafes/ foresight that they had no means to stop it safely.


Perhaps, but trains and trucks do tend to go out of control even with drivers.
Diesel engines can get stuck on, brakes fail.
It's why they have derail points to start with.

Have to see if it was a control system malfunction or a more regular failure which even a driver couldn't have stopped at his controls.

Scott_C
Posts: 735
Joined: Fri Jan 10, 2014 10:49 am
Location: Perth, WA

Re: Autonomous cars? I think not

Postby Scott_C » Fri Sep 21, 2018 12:56 pm

redsonic wrote:Devonport authorities deliberately derailed an out of control driverless freight train in order to stop it. 2 pedestrians were injured:

ABC News

Shows a lack of failsafes/ foresight that they had no means to stop it safely.

I am 95% sure "driverless" in this context means a train that should be controlled by a driver that has runaway without anybody on board, not a train designed to be autonomous (I know the police press release says "automated freight train" but I think this is a miscommunication).

I work in the rail signalling industry (working on an automatic train control project proposal at the moment) and haven't heard anything about TasRail having a GOA4 (autonomous) train operating system. To the best of my knowledge they don't even have speed limit or limit of authority enforcement (GOA1).

User avatar
bychosis
Posts: 5706
Joined: Sat Jan 14, 2012 1:10 pm
Location: Lake Macquarie

Re: Autonomous cars? I think not

Postby bychosis » Fri Sep 21, 2018 1:18 pm

I read an interesting article yesterday. The premise was that autonomous systems are often built to overcome the failings of human inputs, but they can create new unexpected failures.

Automating removes a lot of the human element of decision making. In many cases this is a good thing, but the humans monitoring the automated then rely on the automation and often cannot accept that the system is failing because ‘all the lights are still flashing’. Humans are way more capable than machines at interpreting subtle differences in readings and interpreting them according to otherwise unrelated inputs.
bychosis (bahy-koh-sis): A mental disorder of delusions indicating impaired contact with a reality of no bicycles.

Jmuzz
Posts: 567
Joined: Mon Jan 29, 2018 1:42 pm

Re: Autonomous cars? I think not

Postby Jmuzz » Fri Sep 21, 2018 2:12 pm

Scott_C wrote:I am 95% sure "driverless" in this context means a train that should be controlled by a driver that has runaway without anybody on board, not a train designed to be autonomous (I know the police press release says "automated freight train" but I think this is a miscommunication).


Yeah it was TR11of the TR Class
https://en.m.wikipedia.org/wiki/TasRail_TR_class

No mention of those being remote operated and Tasmania doesn't seem like the sort of place to try and cut jobs with automation.

Some media seem to be running with it though and now changing Driverless to Automated to try and get even more offtrack than the train is.

User avatar
redsonic
Posts: 1082
Joined: Wed Sep 26, 2012 8:08 pm
Location: Brisbane

Re: Autonomous cars? I think not

Postby redsonic » Tue Sep 25, 2018 9:54 am

An article in The Conversation about inattention of the driver in semi autonomous vehicles. Also talks about prioritising the "attention" (computational resources) of the car. I really do not want to see semi autonomous vehicles on the road. I think it is clear that people are hopeless at maintaining alertness when they are only required to monitor things.

Jmuzz
Posts: 567
Joined: Mon Jan 29, 2018 1:42 pm

Re: Autonomous cars? I think not

Postby Jmuzz » Tue Sep 25, 2018 12:16 pm

redsonic wrote:I think it is clear that people are hopeless at maintaining alertness when they are only required to monitor things.


People are pretty hopeless regardless.

I think a more sensible approach is that driverless cars shouldn't be allowed until they are truely driverless and have proven that ability to exceede humans.

Of course they need road trials, and those need to be managed better than the Uber case where the driver was supposed to retain primary control and alertness not play on phone.
You can't pay minimum wage to supervise, it has to be treated as a serious job and pay enough to attract people who will take it seriously.

RobertL
Posts: 730
Joined: Wed Feb 17, 2016 3:08 pm
Location: Brisbane

Re: Autonomous cars? I think not

Postby RobertL » Tue Sep 25, 2018 2:45 pm

Jmuzz wrote:You can't pay minimum wage to supervise, it has to be treated as a serious job and pay enough to attract people who will take it seriously.


Given the enormous cost of developing a driverless car, the amount paid to the human supervisor is a tiny, tiny fraction. So it's one area where it does not make sense to skimp on costs.
Image

User avatar
Ross
Posts: 5380
Joined: Sat Nov 07, 2009 8:53 pm

Re: Autonomous cars? I think not

Postby Ross » Sat Oct 13, 2018 9:39 am

Private autonomous vehicles would be a disaster, comprehensive research finds

https://mobilitylab.org/2018/10/10/priv ... rch-finds/

Who is online

Users browsing this forum: Majestic-12 [Bot]