Brexit, for once some facts.

[PeterL]

[QUOTE="flecc, post: 406269, member: 3"

The biggest problems are and will remain for a long time the legal aspect. Accidents will inevitably happen and who do we accuse and claim from. We can't cross examine cars in courts, do we sue the makers, the suppliers, those who type approved, the maintainers. It's not a simple problem, the more closely it's examined the more difficulties are exposed.

And until these problems are solved not a single driverless car will be on the public streets without a human driver with overriding control.
.
.[/QUOTE]
How about on-board cameras? They are admissible in court as evidence, indeed they often end up incriminating the person in the car at the moment.

 

[PeterL]

Woosh .. I think that that is both understating the current status of sensor technology, the ability of software and exaggerated belief in the set of skills needed to do the tasks. A colleague of mine had robotic assisted surgery on his prostate some 5 years ago... The placement of those knives ,and the propensity to do serious damages to items we hold dear, will dwarf any similar concerns about putting concrete blocks in a row.
More realistic is that robotic technologies make it feasible to use different forms. .. the 3d printing of the house in Nantes a case in point. .. bricks and blocks are the size they are due to the limitations of the human body...

Correct. 3D printing is indeed incredible and again moving at some speed. The example given by BAE of the bicycle repair shop was rather good - I'm sure WOOSH will remember that with some trepidation?

 

[Woosh]

Correct. 3D printing is indeed incredible and again moving at some speed. The example given by BAE of the bicycle repair shop was rather good - I'm sure WOOSH will remember that with some trepidation?

I've seen the clip with the robot that 3D printed the wall.
No match for a human bricklayer.
Same thing with the robot that picked grapes from the vines.
There are always low hanging fruits in any repetitive tasks, especially in agriculture but advances in robotics and artificial intelligence are a lot harder to achieve.
The example of the ping pong game (shown on BBC click) is about the state of the art in software that recognizes shapes and objects. It takes the Japanese team years to get there and illustrates a combination of specialized hardware and software to recognize a simple spherical object, of fixed dimensions, painted in a special orange colour. Generalize that to recognize weeds in your garden will take a lot more work.

 

[PeterL]

I've seen the clip with the robot that 3D printed the wall.
No match for a human bricklayer.
Same thing with the robot that picked grapes from the vines.
There are always low hanging fruits in any repetitive tasks, especially in agriculture but advances in robotics and artificial intelligence are a lot harder to achieve.
The example of the ping pong game (shown on BBC click) is about the state of the art in software that recognizes shapes and objects. It takes the Japanese team years to get there and illustrates a combination of specialized hardware and software to recognize a simple spherical object, of fixed dimensions, painted in a special orange colour. Generalize that to recognize weeds in your garden will take a lot more work.

Maybe we're missing something here? Surely the aim, in construction, will be to automate most, if not all current tasks. If that means changing the colour, the size or even the shape of the brick - then why not? There's no need to make the automation more difficult that it needs to be. Humans do it their way, robots...

 

[PeterL]

Much the same argument for future repairs. I have no doubt that longevity can and will be integral to such buildings. In much the same way that there are no more TV repair shops... buy a new one.

 

[PeterL]

I've seen the clip with the robot that 3D printed the wall.
No match for a human bricklayer..

Did you see the one that printed a bike? When, later, you needed a spare part it printed that as well.

 

[Woosh]

Maybe we're missing something here? Surely the aim, in construction, will be to automate most, if not all current tasks. If that means changing the colour, the size or even the shape of the brick - then why not? There's no need to make the automation more difficult that it needs to be. Humans do it their way, robots...

The Japanese teams are quite good at exoskeletal bionic add-ons but they are not robots which are machines that can recognize shapes and objects around them, work out (albeit limited) goals for themselves and devise ways to achieve them. They need to be able to learn a bit by themselves.

 

[PeterL]

TThey need to be able to learn a bit by themselves.

That's already been done in spectacular fashion - Google's AlphaGO

 

[flecc]

How about on-board cameras? They are admissible in court as evidence, indeed they often end up incriminating the person in the car at the moment.

I don't understand this post, it doesn't answer mine. The only persons in a driverless car are passengers who almost certainly cannot have caused an accident.
.

 

[PeterL]

I don't understand this post, it doesn't answer mine. The only persons in a driverless car are passengers who almost certainly cannot have caused an accident.
.

Oh dear. It will however, in all probability, point the finger. As for the implied absolute responsibility, I'm not a Lawyer, are you? Fairly sure that our law-makers will be fully aware of the issues, as indeed they will across the world. It is surely a problem that needs a solution and one will be found - might even be a good law first time round - who knows?

 

[flecc]

I have no doubt that longevity can and will be integral to such buildings. ... buy a new one.

Nice idea, but we'll have to learn how to demolish the old buildings first, instead of slapping all manner of preservation orders, listing etc on everything on false grounds of merit.
.

 

[flecc]

Oh dear. It will however, in all probability, point the finger. As for the implied absolute responsibility, I'm not a Lawyer, are you? Fairly sure that our law-makers will be fully aware of the issues, as indeed they will across the world. It is surely a problem that needs a solution and one will be found - might even be a good law first time round - who knows?

I'm quoting the lawyers, not my views but theirs.
.

 

[PeterL]

Nice idea, but we'll have to learn how to demolish the old buildings first, instead of slapping all manner of preservation orders, listing etc on everything on false grounds of merit.
.

This is such a negative thread, no matter the subject too many Grumpy Old Men (old farts) on it. Not you of course - just the others FLECC.

 

[PeterL]

I'm quoting the lawyers, not my views but theirs.
.

You never attributed ownership so one assumed it was your opinion. At £500+ per hour no wonder they see problems. If not, they will make them up.

 

[flecc]

as we both know a plane can easily take off, fly to its destination and safely land - all on its own, in any weather.

Not the last three words. Severe cross winds on landing are far beyond automated landing and only coped with by pilot skill.

Much the same with autonomous cars and lorries, I can certainly see the concerns, in the short-term addressed by having convoys of lorries on motorways. Cars much the same with full autonomy easily achieved on a motorway.

Convoy fashion locked to each other on long distances, certainly. We could have that now.

In current urban environments driverless is a very long way off though. Apart from legal difficulties and less due to technological problems than the other elements: The instant actions of impulsive and stupid pedestrians, ditto cyclists, signs, bins, hoardings etc. suddenly blown onto a road by gust of wind, scaffolding or lumps of buildings etc. falling into a road, sudden subsidence making a hole in the road, water, gas and electricity mains suddenly bursting/exploding. Software to deal with all that lot is far beyond possibility

It all happens here in London, more often that most might think.

And of course the ethical issue that the lawyers also worry about. Run over the person suddenly stepping into a vehicle's path or swerve and possibly hit a bus queue of people? Swerve for an animal and risk human life? Software to recognise any animal from any angle and distinguish it from any human form doesn't exist and probably never will.
.

 

[PeterL]

Not the last the words. Severe cross winds on landing are far beyond automated landing and only coped with by pilot skill.

Much the same with autonomous cars and lorries, I can certainly see the concerns, in the short-term addressed by having convoys of lorries on motorways. Cars much the same with full autonomy easily achieved on a motorway.

Convoy fashion locked to each other on long distances, certainly. We could have that now.

In current urban environments driverless is a very long way off though. Apart from legal difficulties and less due to technological problems than the other elements: The instant actions of impulsive and stupid pedestrians, ditto cyclists, signs, bins, hoardings etc. suddenly blown onto a road but gust of wind, scaffolding or lumps of buildings etc. falling into a road, sudden subsidence making a hole in the road, water, gas and electricity mains suddenly bursting/exploding. Software to deal with all that lot is far beyond possibility

It all happens here in London, more often that most might think.

And of course the ethical issue that the lawyers also worry about. Run over the person suddenly stepping into a vehicle's path or swerve and possibly hit a bus queue of people? Swerve for an animal and risk human life? Software to recognise any animal from any angle and distinguish it from any human form doesn't exist and probably never will.
.
.[/QUOTE]
Give a man enough rope...

I'm fairly sure, but not certain, that there is so much conjecture and speculation in this reply that it defies an answer. Do I take it that your last job was that of a H&S Officer?

 

[flecc]

I'm fairly sure, but not certain, that there is so much conjecture and speculation in this reply that it defies an answer. Do I take it that your last job was that of a H&S Officer?

No, not conjecture but as I posted, actual and often frequent events that driverless will have to deal with in urban environments.

H & S officer? That's quite funny to me, I hate the often unnecessary H & S interventions. But this is not a H & S special dangers subject, it's about the possibility of software dealing with wide ranging routine events.
.

 

[PeterL]

No, not conjecture but as I posted, actual and often frequent events that driverless will have to deal with in urban environments.

H & S officer? That's quite funny to me, I hate the often unnecessary H & S interventions. But this is not a H & S special dangers subject, it's about the possibility of software dealing with wide ranging routine events.
.

In some ways you could always compare the reactions of your typical driver to that of the autominous car in such situations. Some of which are frankly out of this world.

I do subscribe to Murphy's Law though. If it can happen, it probably will.

 

[flecc]

in such situations. Some of which are frankly out of this world.

Not out of this world, it seems you'd be surprised by how often some of them occur. Perhaps you'd like to suggest one of the things I've listed that you think unlikely?
.

 

[Woosh]

A colleague of mine had robotic assisted surgery on his prostate some 5 years ago... The placement of those knives ,and the propensity to do serious damages to items we hold dear, will dwarf any similar concerns about putting concrete blocks in a row.

Back in the 90s, I used Cadstar auto router software for PCB layout. Cadstar is not very different to your example of surgery but it is not robotic. A fair degree of human input is still required to finish the work, same way one uses an in car satnav. What would be a surgeon robot is the one that accesses the MRI scan of your friend's prostate then decides a treatment plan all by itself without the need for a human surgeon, anaesthetist, oncologist etc.