Skyscraper City Forum banner
1 - 20 of 47 Posts

·
Registered
Joined
·
6,996 Posts
Discussion Starter · #1 ·



http://www.ted.com/talks/sebastian_thrun_google_s_driverless_car


and today:


Google's driverless car: no steering wheel, two seats, 25mph




Google has demonstrated its own driverless car, a design that does away with all conventional controls including the steering wheel, and says it will build 100 of the vehicles for testing with the eventual aim of "bringing this technology to the world safely".

The company had for several years been testing everyday cars equipped with sensors, navigation equipment and computers to drive themselves but in the meantime it has secretly developed a prototype from scratch that will have no facility for a human to take control, other than an emergency stop button.

An initial 100 testbed versions would retain manual controls, Google said as it unveiled the car on Tuesday. The controls are needed to comply with the law in California which along with Nevada and Florida allows autonomous vehicles but only if a driver can take charge.​


http://www.theguardian.com/technology/2014/may/28/google-reveals-driverless-car-prototype

.
 

·
Registered
Joined
·
1,410 Posts
I can see these really taking off, especially if they can make trains of them, or better still just drive them onto a train and off seemlessly.
 

·
Trains > Cars
Joined
·
470 Posts
Really nice concept, seems to work well. Hopefully more development and public support for technology like this can spur a road-based transportation revolution that highways and roads desperately need.
 

·
Registered
Joined
·
6,996 Posts
Discussion Starter · #4 ·


Clive Sinclair tried to do it in the 1980s, but the ZX81 computer was underpowered for the task.

.
 

·
Registered
Joined
·
6,996 Posts
Discussion Starter · #6 · (Edited)
:)

And the fact it was underpowered, dangerous and only just fit for purpose, like the rest of his cheap tat!
And at a time when Mrs Thatcher avoided any committment to modernising British industry with R&D investment, and wasted it on unemployment benefit. Unlike Norway.


An under-reported fact is that Vince Cable and the Coalition, although cutting infrastructure investment dramatically, has been quietly putting cash into industry, although perhaps not driverless cars.




But will the existing motor manufacturers compete with Google?


There's this:



"MIRA technology park launches academy"

http://www.bbc.co.uk/news/uk-england-coventry-warwickshire-22164080


which has produced this:


Caption:
"MIRA's driverless car: Brilliant British engineering. Gov investing £7.4m in the site"

.
 

·
Registered
Joined
·
1,410 Posts
And at a time when Mrs Thatcher avoided any committment to modernising British industry with R&D investment, and wasted it on unemployment benefit. Unlike Norway.


An under-reported fact is that Vince Cable and the Coalition, although cutting infrastructure investment dramatically, has been quietly putting cash into industry, although perhaps not driverless cars.




But will the existing motor manufacturers compete with Google?


There's this:



"MIRA technology park launches academy"

http://www.bbc.co.uk/news/uk-england-coventry-warwickshire-22164080


which has produced this:


Caption:
"MIRA's driverless car: Brilliant British engineering. Gov investing £7.4m in the site"

.
I know it well, just round the corner from the triumph factory, I wonder if auto pilot motorbikes are next??


Anyway regardless of if it comes from google or home grown tech (and I fancy home grown would better suit our rules and traditions) I think these cars are a better solution for traffic for the cities than any other idea currently out there.

While burying the roads in a great big hole under south London is tempting and might make traveling anywhere south of the river and north of Croydon a lot less painful its unlikely.

I have concerns regarding the liability in the inevitable accident, personally I will never use one (not least because they have 2 too many wheels) but they would be perfect for everyone else :)
 

·
Trains > Cars
Joined
·
470 Posts
The home grown version would probably not store our data forever to sell off to advertisers.
 

·
Registered
Joined
·
7,522 Posts
I can see these really taking off, especially if they can make trains of them, or better still just drive them onto a train and off seemlessly.
Why would you want to drive them on and off a train - it undermines the whole concept?

Nobody will own a driverless car; they are taxi-replacements, not private-car replacements; and this prototype shows exactly why. Every driverless car will offer exactly the same experience, and no driverless car will go faster than any other driverless car. So there is no value in owning your own; you will hire them by the mile.

So, you will call up a driverless car to take you to the station; catch a train to the nearest town to your destination; and another driverless car will be waiting to pick you up and take you to journey's end. Meanwhile the first driverless car will have plugged itself into a recharger, ready for the next hire. Solves all the problems of electric car technology in one go; doesn't matter about limited range, limited top speed, or long recharging. Takes full advantage of short dimensions, smooth ride, short-speed acceleration, and low emissions.
 

·
Trains > Cars
Joined
·
470 Posts
Nerd speaks of a sustainable transport Utopia, a similar vision to one I've had before. We must unite to make this happen.
 

·
Registered
Joined
·
1,410 Posts
Why would you want to drive them on and off a train - it undermines the whole concept?

Nobody will own a driverless car; they are taxi-replacements, not private-car replacements; and this prototype shows exactly why. Every driverless car will offer exactly the same experience, and no driverless car will go faster than any other driverless car. So there is no value in owning your own; you will hire them by the mile.

So, you will call up a driverless car to take you to the station; catch a train to the nearest town to your destination; and another driverless car will be waiting to pick you up and take you to journey's end. Meanwhile the first driverless car will have plugged itself into a recharger, ready for the next hire. Solves all the problems of electric car technology in one go; doesn't matter about limited range, limited top speed, or long recharging. Takes full advantage of short dimensions, smooth ride, short-speed acceleration, and low emissions.
Thats certainly a viable model within built up areas such as London these will wipe out Taxis within a matter of years. If you live in such an area then you can easily get a car whenever you need one, think Boris Bikes are on speed. In fact I could see a scenario were these (ie communal, electric/non polluting) are the only vehicles allowed in central London. Traffic congestion would be solved quite easily, especially were currently drivers block yellow box junctions and swop lanes randomly making holdups for others.

There are some other use cases I am unsure you have thought about:

Do you really see a besuited businessman who drives a 7 series swopping that for one of these? It has no benefit over a train and a taxi. In reality he (well I then :) ) will want a fully fitted mobile office, I will want a place for my laptop, my screens and all my other little bits. I am not going to get into a peon mobile for a 2 hour journey! Once I have my mobile office though I do not give a monkeys if it drives on the road, rolls itself onto the back of a train or is lifted into the sky by elephants riding on a turtle.

In a similar vein we live down a track that is ok for 10 or 11 months of the year. However it has a habit of flooding (to around 18 inches) so we have a car suitable for that. There is no way one of these google mobiles would ever deal with the depth and speed of flow across the track. We would end washed into the Medway and that would be bad. In fact I doubt the autonomous vehicle would even drive down our track, we are not on google street view (although the track is on google maps).

If you were to go shopping and want to park the vehicle in your drive, as opposed to just dropping you off how would that work unless you taught your vehicle about your drive? From what I have read these have no user interaction at all so presumably will rely on data from mapping software, so unless we map our drives the only alternative is to teach our personal vehicles?

What about the people who travel across Europe to places that may not permit these vehicles. In this case they might prefer a vehicle with manual override and controls. Can you imagine how complex loading an autonomous car onto a ferry or the channel tunnel?

In effect we will see a range of vehicles in much the same way as we do now. There will be the taxi replacements (probably similar to the google example above). There will be more traditional cars without controls for personal use and cars with controls that have this as an option.
 

·
Registered
Joined
·
7,522 Posts
Nerd speaks of a sustainable transport Utopia, a similar vision to one I've had before. We must unite to make this happen.
Nothing utopian about it; its the logic of the technology and environmental imperatives, and already entirely feasible. Obviously the taxi and private hire-car industry will try to prevent it; but even so.

Two further aspects need to be developed; which are not apparent in these demonstration prototypes, but may be validly inferred from them;

- driverless cars can function, and must be able to function, in mixed traffic with driven vehicles - buses, trams, delivery vans, public service vehicles, bicycles, motorcycles These will remain driver-controlled. But the full advantage of driverless travel will only be achieved on driverless roads; a seperate network of radial routes from which non-driverless traffic is banned. Damo talks of 'making trains' of driverless vehicles, but that is entirely unneccesssary, as the technology will enable driverless vehicles to tailgate one another closely without any physical linkage. But they cannot similarly tailgate a driven vehicle (as they must leave a prudent margin for driver error in the driven vehicle), and that all those 'prudent margins' add up to far more roadspace than is available without major route construction. Which implies driverless vehicles as a predominantly urban technology, and one that functions best in urban areas with high density occupation. You will use a driverless vehicle for travel within towns; for travel outside towns, and for travel between towns, other modes of transport will continue to be better.

- driverless car technology also implies driverless people-carriers. Because vehicles will be hired by the trip, there will be a need for a range of capacities up to minibus size. Most trips only need two seats (as in the prototype) but other uses (school runs for example) will need four, six or eight; with space for buggies trolleys. In public transport terms, this implies the development of driverless shared taxis (as per the jitney or dolmus), taking over much of the function of buses at the bottom end of the market. Hence, the prospective traveller can use their phone either to call up a single-occupant driverless vehicle (specifying the number of seats required), or they may (at lesser cost) look to book for a shared ride on a jitney. The single occupant vehicle will pick you up at your door, and deliver you to your specified destination anywhere within the urban zone. The jitney will only run on the driverless roads, so it will pick you up and drop you down at points along these designated routes.


One thing tha stands in the way of driveless car adoption at present is the current seriouis disjunction between the imperative of the technology - towards small single user vehicles, ruinning within urban areas and hired by the trip; and the marketing of the project funders- which still suggests driverless technology as somehthing you add to private car ownership. Inherently driverless technology supports low-end public transport functions; but the sponsors have often presented it as providing a future for high-end private transport.
 

·
Registered
Joined
·
7,522 Posts
Thats certainly a viable model within built up areas such as London these will wipe out Taxis within a matter of years. If you live in such an area then you can easily get a car whenever you need one, think Boris Bikes are on speed. In fact I could see a scenario were these (ie communal, electric/non polluting) are the only vehicles allowed in central London. Traffic congestion would be solved quite easily, especially were currently drivers block yellow box junctions and swop lanes randomly making holdups for others.

There are some other use cases I am unsure you have thought about:

Do you really see a besuited businessman who drives a 7 series swopping that for one of these? It has no benefit over a train and a taxi. In reality he (well I then :) ) will want a fully fitted mobile office, I will want a place for my laptop, my screens and all my other little bits. I am not going to get into a peon mobile for a 2 hour journey! Once I have my mobile office though I do not give a monkeys if it drives on the road, rolls itself onto the back of a train or is lifted into the sky by elephants riding on a turtle.

In a similar vein we live down a track that is ok for 10 or 11 months of the year. However it has a habit of flooding (to around 18 inches) so we have a car suitable for that. There is no way one of these google mobiles would ever deal with the depth and speed of flow across the track. We would end washed into the Medway and that would be bad. In fact I doubt the autonomous vehicle would even drive down our track, we are not on google street view (although the track is on google maps).

If you were to go shopping and want to park the vehicle in your drive, as opposed to just dropping you off how would that work unless you taught your vehicle about your drive? From what I have read these have no user interaction at all so presumably will rely on data from mapping software, so unless we map our drives the only alternative is to teach our personal vehicles?

What about the people who travel across Europe to places that may not permit these vehicles. In this case they might prefer a vehicle with manual override and controls. Can you imagine how complex loading an autonomous car onto a ferry or the channel tunnel?

In effect we will see a range of vehicles in much the same way as we do now. There will be the taxi replacements (probably similar to the google example above). There will be more traditional cars without controls for personal use and cars with controls that have this as an option.
I have thought a great deal about this; indeed posted a lot on this forum albiet a year or more ago.

Essentially driverless technology prsents a viable solution to a problem; but as yet the publc (and commercial market) are unclear as to which exactly which is the problem it is best suited to be a solution.

For simplicities sake we can divide these problems into two clusters:

Cluster One - driverless technology as a solution to the problem of decreasing private car functionality. You can't drive a car while digitally connected; and private car makers and digital providers have tended to see the driverless car as a means by which car travellers remain connected while on the move.

Cluster Two - driverless techology as a solution to the problem of urban road capacity. If you can get urban traffic to move faster and closer together, then the existing road network can carry vastly more people than currently. Furthermore, if there is no driver, vehicles can be much shorter and use less space.

As you will have gathered, the current promoters of the technology are now mainly interested in cluster one; but my contention is that the potential capabilities of this have been unrealistically talked-up. The arguments you make exactly illustrates this point. Cluster two is where the technology offers major added value - so much that, in my viewe, the imperative of the technology wll determine that this is the way things go.

Thinking in general terms.

- the idea that driving yourself to work is an aspect of high social status is very much a late 20th century fashion. If you look back before WWII, there is a clear distinction; bank managers, solicitors, commercial travellers and country doctors drove their own cars; bank directors, judges, chief executives, landed gentry and hospital consultants travelled within town by taxi, and outside town by first-class rail travel. The really wealthy might have a chauffeured Rolls to take them from their country seat to the station, but the basic principle remained. What the happened wias the War, as a consequence of which, trains got slower and road vehicles got faster. Moreover, the Luftwaffe and RAF solved a key disadvantage of the private car - that of finding a place to park it your destination. But the technolgical wheel has taken a further turn; trains are now getting faster by the year, and cars are (in practice) getting slower. I used to do a regular trip from Manchester to Leeds by road in 45 minutes; now it would take me over 90, if I drove (which is why I no longer do). Driverless technology won't solve this, it allows driverless cars to travel safely at the same speed in the same lane; it won't allow them to overtake one another - though it may well support differntial speed lanes on driverless roads.

- what we already see is the private car morphing from a work-related vehicle driven by men; to a leisure and family-related vehicle driven by women. In my view, driverless tecnhology will accelerate this trend; people who can afford private cars will continue to keep them for week-end use, for shopping, and to ferry the kids around. But they won't use them for getting to work - unless they live well outside town, and the tecnology will tend to make towns much more attractive. So it may not make much difference to more affluent populations; but it will be a big incease in transport functinality at the bottom end of the income scale. The populations with far the largest taxi use in the UK is in Liverpool, where it provides the standard means for low income famillies to get around cheaply, in neighbourhoods that are under-served by public transport and where private cars have a very limited life expectancy. The driverless car will be Kirkby writ large.
 

·
Registered
Joined
·
1,410 Posts
Nothing utopian about it; its the logic of the technology and environmental imperatives, and already entirely feasible. Obviously the taxi and private hire-car industry will try to prevent it; but even so.

Two further aspects need to be developed; which are not apparent in these demonstration prototypes, but may be validly inferred from them;

- driverless cars can function, and must be able to function, in mixed traffic with driven vehicles - buses, trams, delivery vans, public service vehicles, bicycles, motorcycles These will remain driver-controlled. But the full advantage of driverless travel will only be achieved on driverless roads; a seperate network of radial routes from which non-driverless traffic is banned. Damo talks of 'making trains' of driverless vehicles, but that is entirely unneccesssary, as the technology will enable driverless vehicles to tailgate one another closely without any physical linkage. But they cannot similarly tailgate a driven vehicle (as they must leave a prudent margin for driver error in the driven vehicle), and that all those 'prudent margins' add up to far more roadspace than is available without major route construction. Which implies driverless vehicles as a predominantly urban technology, and one that functions best in urban areas with high density occupation. You will use a driverless vehicle for travel within towns; for travel outside towns, and for travel between towns, other modes of transport will continue to be better.
A number of things, first why must public service vehicles be manually controlled? It makes more sense for them to be automated too so all the vehicles work together. For example a bus wishing to cross a busy junction could be allowed to by automatic vehicles if they talk to each other. As it stands these vehicles dont talk to each other but I imagine its only a matter of time, especially in the model where they replace taxis and public service vehicles.

Secondly these clearly work in mixed traffic as Google are driving them around California already.

Surely the gap between an automated and non automated vehicle, assuming no inter vehicle communication will be the same? ie the distance the automated vehicle needs to stop.

You are not going to get driverless radial roads, in London at least, for a long time. You may find the congestion charge zone allows for these to operate for free or some such tweak to encourage them from that perspective. Clearly we are unlikely to build roads in the city so that just leaves using existing road infrastructure. There are very few roads that would support automated only lanes, just look at the longest stretches of bus lane, criss crossed with side roads, pedestrian crossings etc.

- driverless car technology also implies driverless people-carriers. Because vehicles will be hired by the trip, there will be a need for a range of capacities up to minibus size. Most trips only need two seats (as in the prototype) but other uses (school runs for example) will need four, six or eight; with space for buggies trolleys. In public transport terms, this implies the development of driverless shared taxis (as per the jitney or dolmus), taking over much of the function of buses at the bottom end of the market. Hence, the prospective traveller can use their phone either to call up a single-occupant driverless vehicle (specifying the number of seats required), or they may (at lesser cost) look to book for a shared ride on a jitney. The single occupant vehicle will pick you up at your door, and deliver you to your specified destination anywhere within the urban zone. The jitney will only run on the driverless roads, so it will pick you up and drop you down at points along these designated routes.
Have you done a school run or even had any dealings with people with buggies? The buggies (and related child paraphernalia) live in the boot of your car. If you need the buggy to walk to the shops you get it out, otherwise its always in the car so when its needed you have it. Even if you have space in your house for buggys etc most parents I know leave them in their car so they don't clutter up their hall.

If you do the school run in one of these what happens when you get to the school? If you need to drop one child at one school and another at a different school you need to leave your stuff in the car while you walk them through the gates. Then come back to the same car to do the other school. This is going to make the school run worse not better.

A hire car for a school run is never going to happen, at least replacing people who currently use their own vehicles, its just not practical.

Private ownership will remain because of these reasons and a huge range of other reasons people currently own their own cars.

One thing tha stands in the way of driveless car adoption at present is the current seriouis disjunction between the imperative of the technology - towards small single user vehicles, ruinning within urban areas and hired by the trip; and the marketing of the project funders- which still suggests driverless technology as somehthing you add to private car ownership. Inherently driverless technology supports low-end public transport functions; but the sponsors have often presented it as providing a future for high-end private transport.
Why is it imperative that this technology be fitted to small vehicles in urban areas?

It clearly supports both cases. And people will want both. The technology does not lean towards any particular vehicle at all (in fact Google developed it originally on off road trucks and 4x4s).

Overall think your use cases through better, not what you want people to do or how you want people to use this technology but how people currently use their resources and how to make their resources do more for less. I get the impression you are trying to make this technology fit your vision of the future.
 

·
Registered
Joined
·
7,522 Posts
A number of things, first why must public service vehicles be manually controlled? It makes more sense for them to be automated too so all the vehicles work together. For example a bus wishing to cross a busy junction could be allowed to by automatic vehicles if they talk to each other. As it stands these vehicles dont talk to each other but I imagine its only a matter of time, especially in the model where they replace taxis and public service vehicles.
Fire tenders, police cars, ambulances will all need to continue to be driver controlled when in 'blue light' mode; for the simple reason that they will have to continue to work when the automatic system fails. This is related to the fact that driverless vehicles cannot overtake; as there will be no 'gaps' in the forward traffic stream. Which in turn means that driverless control techology will need a 'blue light' override that ensures that vehicle pull in to allow emergency services past. Of course, in time, all such vehcles may be expected to have switchable driverless technology for standard mode travel.

Secondly these clearly work in mixed traffic as Google are driving them around California already.

Surely the gap between an automated and non automated vehicle, assuming no inter vehicle communication will be the same? ie the distance the automated vehicle needs to stop.
Certainly the vehicles can be and are used in mixed traffic, but only at the cost of consuming more road space than driven vehicles, to allow for driver stupidity. Which is where your second point is exactly wrong. When we drive on highly congested road (e.g motorways) we are now forced to assume a degree on non-stupidity in other road users. Which allows such roads to have much higher vehicle capacity than they were originally designed for; because vehicles travel closer together at higher speed. But there is a limit to the bunching that can be achieved in driven cars; when tailgating gets too close, or due to overtaking, capacity falls. Driverless technology provides a step increase in capacity by allowing vehicles to travel safely much more closely separated than their theoretical stopping distance. And not having overtaking adds another big capacity gain. And that probably will imply inter-vehicle communication.

Remember that at urban speeds of 20-30 mph; 'thinking distance' is as long a 'breaking distance'. The driverless care thinks instantaneously; but it has to assume a 'thinking distance' allowance or nearby driven cars. But if the nearby car is itself driverless (and both cars know this) no such "thinking distance" allowance is necessary.

You are not going to get driverless radial roads, in London at least, for a long time. You may find the congestion charge zone allows for these to operate for free or some such tweak to encourage them from that perspective. Clearly we are unlikely to build roads in the city so that just leaves using existing road infrastructure. There are very few roads that would support automated only lanes, just look at the longest stretches of bus lane, criss crossed with side roads, pedestrian crossings etc.
Do the numbers. An urban main road with junctions at grade has a peak capcity of around 1,000 persons in private cars per hour per traffic lane in each direction - assuming 1.2 persons per private car. If that is upgraded to an urban motorway with grade separated junctions,and so long as traffic can keep moving along it at at least 40 mph, then capacity increases to 2,000 persons per hour per lane. If we assume vehicles of the sort in the prototype, then they will consume half the roadspace each, and travel with half the distance between them. That gives at least 4,000 persons per lane per hour in each direction along driverless-only roads. Converting an existing radial route to driverless only operation will give roughly the same capacity increase as building a completely new urban motorway.

So there won't be a problem finding the space for driverless-only lanes; there will be more of a problem picking up the weeds on the residual non-driverless lanes.

Have you done a school run or even had any dealings with people with buggies? The buggies (and related child paraphernalia) live in the boot of your car. If you need the buggy to walk to the shops you get it out, otherwise its always in the car so when its needed you have it. Even if you have space in your house for buggys etc most parents I know leave them in their car so they don't clutter up their hall.

If you do the school run in one of these what happens when you get to the school? If you need to drop one child at one school and another at a different school you need to leave your stuff in the car while you walk them through the gates. Then come back to the same car to do the other school. This is going to make the school run worse not better.

A hire car for a school run is never going to happen, at least replacing people who currently use their own vehicles, its just not practical.

Private ownership will remain because of these reasons and a huge range of other reasons people currently own their own cars.
Agree with all that damo (except the bit about the school run). I suspect our post crossed over, but I emphasise that I do not see the end of the private care for family use, only for commuter use. But there will be many more people who don't own any car (especially in urban areas) becuse driverless travel will be more convenent. They will hire driverless cars for the school run.

When I was finishing school, it was a matter of course that I learned to drive before going to uni. Almost all my friends did the same. And I bought my first car some three years before I bought my first flat. Now it is the other way round. The median age for learning to drive is now in the late 20s and gets later each year that passes; and most yong people buy a flat long before they buy their own car. Wheras it used to be the case that young people bought a car when they got their first real job; now they buy a car when they get their first child. In the future, I see many families never needing a car; simply because they have never had to get round to learning to drive.

Why is it imperative that this technology be fitted to small vehicles in urban areas?

It clearly supports both cases. And people will want both. The technology does not lean towards any particular vehicle at all (in fact Google developed it originally on off road trucks and 4x4s).

Overall think your use cases through better, not what you want people to do or how you want people to use this technology but how people currently use their resources and how to make their resources do more for less. I get the impression you are trying to make this technology fit your vision of the future.
Again, do the numbers. 90% of the cost of a modern car is devoted to improving 'driver experience'. Giving the driver leg-room, providing access to the controls, directing them with air-conditioning and in-car entertainment, keeping them alert, making sure they can see. With driverless cars there is no driver experience; and none of this is needed. So they can be much, much cheaper and smaller.

And in the market; cheaper and smaller usually wins.
 

·
Not Cwite There
Joined
·
9,242 Posts
I'm not sure that vehicle separation can ever be below stopping distance - automatic railway systems still work on the basis of stopping distance + margin, so I don't see why it would be different for roads. Given that people often keep less than stopping distance on motorways I suspect driverless vehicles might actually reduce to a capacity reduction as stopping distances would be enforced by computers.
 

·
Registered
Joined
·
9,917 Posts
Have you done a school run or even had any dealings with people with buggies? The buggies (and related child paraphernalia) live in the boot of your car. If you need the buggy to walk to the shops you get it out, otherwise its always in the car so when its needed you have it. Even if you have space in your house for buggys etc most parents I know leave them in their car so they don't clutter up their hall.

If you do the school run in one of these what happens when you get to the school? If you need to drop one child at one school and another at a different school you need to leave your stuff in the car while you walk them through the gates. Then come back to the same car to do the other school. This is going to make the school run worse not better.

A hire car for a school run is never going to happen, at least replacing people who currently use their own vehicles, its just not practical.

Private ownership will remain because of these reasons and a huge range of other reasons people currently own their own cars.
I can vouch that all of this is very true. Going through it at the moment, all this kerfuffle dealing with kids & their kit is serious grief - even when you leave the buggies, toys and car seats in your vehicle. The thought of pulling this stuff in and out twice every day and finding extra room in the house to store them, while looking after kids is not tenable. And that's before you start thinking how many people view their cars as private/personal space.... Public vehicles will only appeal to a certain percentage of motorists.
 

·
Registered
Joined
·
7,522 Posts
I'm not sure that vehicle separation can ever be below stopping distance - automatic railway systems still work on the basis of stopping distance + margin, so I don't see why it would be different for roads. Given that people often keep less than stopping distance on motorways I suspect driverless vehicles might actually reduce to a capacity reduction as stopping distances would be enforced by computers.
That is true with respect to motorways; but not in respect of urban roads. Remember that the 'stopping distance' comprises both 'thinking distance' and 'braking distance'. At urban road speeds - 20mph to 30mph - the two components are about equal. So, on a driverless-only road, the elimination of the need for 'thinking distance', halves the separation of the vehicles.

Which is why this particualar pilot exercise represents a major step forward. Thus far, Google and its chums have tended to show off driverless technology as an add-on to high-range cars; with the implied message 'spend more on your car to be able to remain connected while travelling". But it does not appear that the economics stacks up (even if we assume substantial public subsidy). Rich people with swanky cars tend to live out of town, and the technolocy requires that driverless cars can only function on roads and tracks that are fully pre-digitised to a high degree of definition. Ad driverless technology offers much less added value on country roads and motorways.

This exercise shows that the economics - certainly to start with - favour driverless technology in high density urban roads. This minimises up-front digitising costs, and plays to the strengths of the technology in low-speed, high density traffic. But that in turn implies a taxi-style vehicle; which in turn makes possible a design that ditches most of the costly driver-specific facilities and space. Which can then be offset against the extra cost of the driverless control kit. Furthermore, if the adoption of driverless technology can generate a massive expansion in urban road capacity - without investment in new roads - then the case can be made that the up-front digitising costs (which are massive, as will be their cotinued maintainance) can be proposed for public funding; and that could be the factor tipping the business-case into the black.

Of course - as damo points out - this is all blue-sky thinking. We don't yet know that the numbers will stack up; we don't yet know that sufficient city dwellers will find the prospect of driverless taxis an attractive alternative. Very few people nowadays use a taxi to commute to work - even in Liverpool. Effectively this technology stands or falls on whether the economics can favour driverless taxis as a viable standard commuting mode for the future. But to demonstrate that, Google needs a fleet of pilot vehicles; not just a few 4wd people carriers with mini-lighthouses on their roofs.
 
1 - 20 of 47 Posts
Top