Can 3 million EVs replace 30 million oil cars?

Piddles

tim_p wrote: »

Quick thought... who maintains, insures, cleans, charges, checks these millions of vehicles? Who buys them in the first place? How does a tiny 3 person pod cope with 3 people going shopping? How would a dog owner manage? A smoker or vaper? Disabled person?
Gonna be some issues come Sat night / Sunday morning, what with chaos left behind by drunk passengers.
Gonna be fights over the setting of the aircon (source: my wife)
Gonna be fights over the choice of music / radio channel.
Gonna be fights over unwashed fellow ‘podders’ (ooh! coined a phrase!)

I know I’m being completely negative here but I’m imagining a time when us plebs are trundling about the place like animals while the rich swan around in their luxury cars, probably in priority lanes too.

Google Insurance: " You appear to be making a claim against one of our cars. Would you like to buy some more underpants?"

tunnel

GreatApe wrote: »

85% of the population live in urban (dense) areas

Source?

GreatApe wrote: »

These will be fine for many rural areas too. Any area where in a square KM there are 100+ households can benefit from these. That's most rural areas too

Are you sure about that cos where i live(rural area) 100 houses per square km there aint, in fact, that would cover a large proportion of Derbyshire, Notts & lincs for a start.

GreatApe wrote: »

Why so gloomy the worse case is we don't go from 30m to 3m but maybe 30m to 10m still a factor of 3x improvement

That would be 3m in the dense urban areas and 7m cluttering our green and pleasant countryside....no thanks

Martyn1981

How about, instead, we just replace all cars with Dodgems? Might not be the most practical solution but it would be a F'n good laugh getting from A to anywhere.

zeupater

Piddles wrote: »

Let's start with Artificial Intelligence. A lot of guff is talked about AI by vested interests and marketing people trying to jump on the bandwagon by dressing up regular algorithms as AI. The anywhere-to-anywhere autonomous vehicle service that most people seem to imagine is a phenomenally difficult task that will require AI way, way more advanced than we have now. That level of AI is probably at least ten years away, quite possibly a fair bit more.

There's been around $100 billion (yes, billion) invested in autonomous vehicles, so it's difficult to imagine those investments not leading anywhere (another investment today). Here in Cambridge we are blessed with not one but two AI autonomous vehicle software businesses five.ai and wayve.ai. They are minnows in comparison to the Waymos of this world (five.ai has received $37.7 million in funding). Their pitch seems to be that European streets are very different to the wide, uncongested, mild-weathered roads of the current Phoenix testing grounds, and are just waiting for the big tech companies to come over and buy them up. The apparent lack of ambition seems a shame.

What we also have in Cambridge are the guided busways. These are disused railways re-purposed as segregated bus routes. There's a £4 billion embryonic plan to grow this into a trackless autonomous tram or smaller autonomous vehicle system including tunnelled lines under the city. The government has provided a £3.2 million grant for autonomous vehicle trails that are due to start on the existing guided busway this year.

The point about these busways are their segregated nature, so you don't need the full AI stack and can work with existing technology. They can then gradually break out into geo-fenced areas in the science parks, hospital campus and the city centre, then villages and towns on the routes as the technology is proved and improved. We have seen individually owned internal combustion engined vehicles being gradually squeezed out of city centres over many, many years and into Park & Rides, etc. This is likely to accelerate as the health effects of high pollution level in the city centres are better understood and become more politically unacceptable. So these autonomous vehicles will gradually fill that void.

One of the disused railways goes onto Oxford (where there's a third AI AV company called Oxbotica), so again that is likely to re-purposed for autonomous vehicles, breaking out into the villages and towns en route. And so it goes on.

From the other side, the pure AI AV companies like the above (or those big US tech companies by then) will later be producing autonomous trucks geo-fenced to motorways and trunk roads, which are likely to be the first AVs you'll see on public roads, not least due the growing shortage of truck drivers (the average age is now 59). As technology and acceptance evolves new vehicle types will gradually make the transition and gradually overlap with urban AVs.

So, in short it's going to be a very gradual, almost imperceptible change, like so many technological advances, as AVs gradually out competes the individually owned vehicle on cost and convenience.

History has a likely parallel. The transition from horse and cart to internal combustion engine took around ten years, with the horse and cart getting banned from many cities in the US on grounds of......pollution. And to put it bluntly, getting in the way.

Hi

Maybe, but the question revolves around the level of traffic flow efficiency gain whilst there's a mix of intelligence levels on the road ... it's almost certain that the infrastructure needs to cope with all of the modes of road transport and that anywhere near 100% AI saturation is nigh on impossible unless seriously strict legislation effectively forces the public to comply: doubt that would go down well at the ballot box, so it's not likely to happen soon, if at all!

Regarding AI ... not quite the same sector or the level of development as we're considering here, but I do have well over 20 years of experience in the field & have often come across the very interaction issues being discussed, that's why I'm normally concerned whenever I hear or see mention of efficiency gains which, although theoretically possible using AI, are effectively unattainable in the real world ... those selling a solution are often tempted to over-promise, often in the knowledge that the recognition of flawed logic is well in the future ....

It's not that AI won't/can't help, it's just that the potential of employing AI needs a little real world logical approach ....

HTH
Z

Piddles

zeupater wrote: »

Maybe, but the question revolves around the level of traffic flow efficiency gain whilst there's a mix of intelligence levels on the road ... it's almost certain that the infrastructure needs to cope with all of the modes of road transport and that anywhere near 100% AI saturation is nigh on impossible unless seriously strict legislation effectively forces the public to comply: doubt that would go down well at the ballot box, so it's not likely to happen soon, if at all!

Part of Uber's pitch to the authorities is it's supposed to ameliorate traffic and reducing personal car ownership. The reality is it that in some US cities it would appear to be doing the opposite (at least on the first point). When you take your Uber you are taking up as much road space as you would be in your own car, plus once Uber has dropped you, they drive around empty to their next fare. Plus as it's so cheap and convenient people are using it more. Fine, introduce UberPool where you share with 2 or 3 other riders going in a similar direction, it proves popular, problem solved. Not so fast. At the lower price point you start getting people off public transport as you are offering a point-to-point (with a couple of stops along the way) service that's faster and more convenient than public transport which never goes directly from where you are to where you want to be. So you end up with even more traffic.

Yeah, I'm agreeing with you.

Back to Cambridge, the fastest growing city in the country and where the road infrastructure has reached capacity (I can vouch for that!), so a good case study. Bringing back these disused railways is added segregated capacity using the latest available technology with a kind of Uber-UberPool supplementing buses with point-to-point 10-15 seat autonomous shuttles (at least in this years trial). The number of points will be initially limited, but will grow as the network expands. The segregation means no congestion, so fast reliable journey times pulling commuters from their cars building volume, reducing prices, building further volume, expanding the network, and so on.

The crucial point is the segregation allows you to reduce congestion before the AI comes fully of age. As it does, these shuttles can break out further and further into the wild and into the utopia of saved lives, inclusion of the excluded, carbon neutral transport, minimal congestion, competitive economy, etc.

The main driver will be price. So demand lead, not government imposed.

tim_p

What’s going to replace the governments revenue from the 27 million vehicles?
VED, VAT, fuel, parking, fines........ we’re not talking small change here but a major source of treasury revenue.

Piddles

tim_p wrote: »

What’s going to replace the governments revenue from the 27 million vehicles?
VED, VAT, fuel, parking, fines........ we’re not talking small change here but a major source of treasury revenue.

You forgot the congestion charge. £250 million for TfL. Coming to a city near you soon....:rotfl:

For most town and cities (not London) around a quarter to a third of real estate is given over to car parks and associated infrastructure no longer required in the age of AVs. Much of it owned by the council. This can be redeveloped for much needed housing and business premises for jobs. Local councils will be big winners with a significant net gain from revenue from property taxes and rents. Removal of congestion and the reduction of transport costs (they'll be using AV services too of course) should mean a big reduction in the costs of council services. In the US, a Morgan Stanley study indicated a net gain of half a trillion dollars for the municipalities. In the UK it ought to reduce central government support for the councils.

For central government, this additional urban housing means less (legislate for no?) urban sprawl and less new transport infrastructure. The near elimination of road deaths and injuries could reduce NHS costs by 15% (sorry, US figures again) with equivalent figures for the police and fire services. Then there's the care and benefits costs (sometimes lifetime) for the seriously injured, many of whom should have been working and contributing taxes.

Similarly, there's the environmental costs to the NHS from transport pollution. AVs are likely to be rebuilt every 3 years due to their high mileage, and their centralised ownership will allow the government to mandate the cleanest technology as it comes available. Now it's more like a 12 year cycle.

The government has earmarked £50 billion to be spent on road improvements. Will any of that be necessary in the age of AVs? In fact, there ought to be a moratorium on all government transport capital projects (and some non transport ones as AVs will transform our lives in ways we haven't yet appreciated).

Transport is a significant element of a county's global competitiveness. A cheaper, more efficient transport system will be a big boost to the economy and tax revenues overall.

You can bet the government will be pushing hard for this. For all our sakes.

tim_p

Road improvements? They need to properly maintain the existing ones before worrying about improvements. Besides these fantasy vehicles will still need decent roads to drive on. Car parks can only be sold off once whereas road users are the gift that keeps giving.

zeupater

Piddles wrote: »

Part of Uber's pitch to the authorities is it's supposed to ameliorate traffic and reducing personal car ownership. The reality is it that in some US cities it would appear to be doing the opposite (at least on the first point). When you take your Uber you are taking up as much road space as you would be in your own car, plus once Uber has dropped you, they drive around empty to their next fare. Plus as it's so cheap and convenient people are using it more. Fine, introduce UberPool where you share with 2 or 3 other riders going in a similar direction, it proves popular, problem solved. Not so fast. At the lower price point you start getting people off public transport as you are offering a point-to-point (with a couple of stops along the way) service that's faster and more convenient than public transport which never goes directly from where you are to where you want to be. So you end up with even more traffic.

Yeah, I'm agreeing with you.

Back to Cambridge, the fastest growing city in the country and where the road infrastructure has reached capacity (I can vouch for that!), so a good case study. Bringing back these disused railways is added segregated capacity using the latest available technology with a kind of Uber-UberPool supplementing buses with point-to-point 10-15 seat autonomous shuttles (at least in this years trial). The number of points will be initially limited, but will grow as the network expands. The segregation means no congestion, so fast reliable journey times pulling commuters from their cars building volume, reducing prices, building further volume, expanding the network, and so on.

The crucial point is the segregation allows you to reduce congestion before the AI comes fully of age. As it does, these shuttles can break out further and further into the wild and into the utopia of saved lives, inclusion of the excluded, carbon neutral transport, minimal congestion, competitive economy, etc.

The main driver will be price. So demand lead, not government imposed.

Hi

Yet the main point in question still revolves around claims of AI enabled autonomous vehicles reducing journey times, traffic bottlenecks etc through eliminating wait times at junctions and maybe even allowing traffic flow control to be handled by AI system intra-vehicular negotiation as opposed to the current stop light system. The theory & often touted efficiency data is heavily dependent on autonomy fully replacing manual control, which it's unlikely to do, resulting in overselling of efficiency/productivity.

Synchronised launch at stop lights will be a reality, there's little doubt about it, however, until all vehicles of all types are AI enabled, then the efficiency improvement for the whole launch will depend on the number of manual vehicles and the position of those vehicles in the queue ... AI based autonomy may bring better reaction times into play, but if the vehicle in front is manually controlled it makes little difference.

In summary, autonomy will help improve efficiencies & may reduce the number of vehicles being built, but the resultant may easily be more vehicle miles travelled, more wear on road surfaces, more load on unavoidable traffic flow pinch nodes and an annoying shortage of available units at peak demand times ... yes, I've seen the attempts to address this issue when posed, but as they normally involve social engineering across all sectors, they're also not too likely!

HTH
Z

Piddles

Conceptually, the autonomous vehicle companies adding their hardware to existing car models isn't helping as it seems to be leading to the impression that we'll be trading in our cars for such vehicles. In Europe and other more densely populated parts of the world, it's more likely to start as a public transport addition/distruptor.

Think existing buses chopped in thirds, taking people from closer to where they are, more directly to closer to where they want to be, departing/arriving at times closer to what they want (none of which existing public transport is very good at) and at a significantly lower cost.

Related links:

Cambridge develop one of the first autonomous shuttle services in the UK

Rethinking Transportation 2020-2030 Quite a long but informative PDF document.

.TaaS (Transport as a Service) will provide 95% of the passenger miles traveled within 10 years of the widespread regulatory approval of AVs. By 2030, individually owned ICE (internal combustion engine) vehicles will still represent 40% of the vehicles in the U.S. vehicle fleet, but they will provide just 5% of passenger miles.

Behavioral issues such as love of driving, fear of new technology or habit
are generally believed to pose initial barriers to consumer uptake. However,
Pre-TaaS companies such as Uber, Lyft and Didi have invested billions of
dollars developing technologies and services to overcome these issues. In
2016, Pre-TaaS companies drove 500,000 passengers per day in New York
City alone.1 That was triple the number of passengers driven the previous
year. The combination of TaaS’s dramatically lower costs compared with
car ownership and exposure to successful peer experience will drive more
widespread usage of the service. Adopting TaaS requires no investment or
lock-in. Consumers can try it with ease and increase usage as their comfort
level increases. Even in suburban and rural areas, where wait times and
cost might be slightly higher, adoption is likely to be more extensive than
generally forecast because of the greater impact of cost savings on lower
incomes. As with any technology disruption, adoption will grow along an
exponential S-curve.

Or if you prefer a video Clean Disruption - Energy & Transportation