View previous topic | View next topic

Do androids dream of electric cars?

Page 1 of 32
Goto page 1, 2, 3 ... 30, 31, 32  Next

Jenny
1312650.  Tue Feb 05, 2019 10:40 am Reply with quote

Well done those who immediately clocked the Philip K Dick reference, but in fact what I would like this thread to be about is the use of AI and technology and its likely practical impact on all our futures.

For example - all the electric cars seem to be currently being tested in California, though I have heard of a small test being conducted here in Portland, Maine. I'm sure they'll get interesting and useful results, but I seriously can't see how the technology in its current state is going to be any use in a climate like ours in northern New England, or indeed in the northern mid-West states.

Our road markings have to be repainted quite frequently as it is, because extensive snowplowing during the winter because of ice and snow on the roads wears them off quite fast. We also get a lot of potholes in the spring because of winter freeze and thaw. And of course during the winter there are times when the roads are covered with snow and slush, and road markings at the side of the road where the plows don't touch get covered up. How is this going to work with a technology that seems to depend on the car being able to 'read' road markings?

And in further thoughts about the practical impact, how are all those electric cars going to be charged in sufficient time for anybody who has to drive more than a limited distance (and back again). For example, it is quite routine here in Maine - and more so in many midwestern states - to drive two or three hours to get to a destination. I have a friend in Rockland and it takes me that long to drive up and see her, and then I drive back the same day.

How are we going to generate the electricity to cover the necessary journeys in a green and sustainable way?

What other practical impacts do QIers think newer technology will have on our future lives? We're already at the stage where internet access is practically a human right, as essential as water or electricity.

 
PDR
1312652.  Tue Feb 05, 2019 11:03 am Reply with quote

I think you've conflated two issues - electric cars and autonomous cars. Electric cars don't have to be autonomous and autonomous cars don't have to be electric.

Electric cars suffer all the issues you raise in respect of where the energy comes from, how they can be charged to be as usable plus the glossed-over issue of the safety aspects of that stored energy. People talk about hydrogen (often for use in fuel cells, but we'd be better off just burning it in the fully matured internal combustion engine technology) but seem vague about where the hydrogen would come from - it doesn't occur naturally in any significant quantities, so it would have to be "made" with all the same issues as finding a way to generate enough electricity.

Autonomous cars, on the other hand, are technically feasible right now. There are decisions to be made about the vehicle vs the infrustructure - do we bury guide wires and junction transponders in our roads, do we provide a wireless information network to inform and guide them, or do we try to make the cars very intelligent so that they don't need these things. There is a continuum of trade-off choices between these two extremes. From a technical viewpoint all of this could be done now - today, or at most by the end of the month.

IMHO the potential bum-bite is legal rather han technical. If an autonomous car crashes into something who is liable for the damage? Is it the occupant, the owner, the manufacturer, the software architect who devised it or the certification authority who allowed it onto the road?

We've already seen instances where autonomous systems which were supposed to be "supervised" by the occupant have killed people becuase the occupant failed to supervise. In one a Tesla "autopilot" killed its owner because the owner would routinely sit in the passenger seat and watch movies, failing to see the car drive itself under a truck. In another an autonomous car or a trial with a dedicated "safety driver" was driving at night on an unlit road when it hit a pedestrian. The pedestrian was crossing the road, but wasn't spotted by the safety driver because the auton9omous car didn't have any need to undip the headlights.

Imagine the interaction of a few dozen incidents like this with a US-style litigation culture. At the moment I really struggle to see any enterprise being able to insure the sale or operation of autonomous cars for this reason. So the problem isn't technical - it's legal/cutural/social IMHO.

Ä0.0000006 supplied,

PDR

 
Leith
1312675.  Tue Feb 05, 2019 2:47 pm Reply with quote

PDR wrote:
So the problem isn't technical - it's legal/cutural/social IMHO.

Bristol has been been a centre for UK autonomous car trials over the past few years, and that's certainly the area they are focusing on round here, with AXA heavily involved in looking at the legality and insurability side, and others carrying out trials to gauge public response to the vehicles.

 
PDR
1312686.  Tue Feb 05, 2019 6:42 pm Reply with quote

My own employer may or may not have developed extensive expertise and experience in the field of autonomous aviation - I can neither confirm nor deny such a thing. But I can state that my employer has explicitly decided NOT to venture into what might appear to be the logical near-neighbour application of such capabilities (if they exist, on which I cannot comment) because Corporate Counsel cannot see how to bound the legal/commercial/cultural risk in any way that might make it a practicable enterprise.

PDR

 
Awitt
1312687.  Tue Feb 05, 2019 7:38 pm Reply with quote

PDR wrote:
Quote:
Ä0.0000006 supplied,


Gone up since your return with inflation?

 
dr.bob
1312716.  Wed Feb 06, 2019 10:07 am Reply with quote

Jenny wrote:
And in further thoughts about the practical impact, how are all those electric cars going to be charged in sufficient time for anybody who has to drive more than a limited distance (and back again). For example, it is quite routine here in Maine - and more so in many midwestern states - to drive two or three hours to get to a destination. I have a friend in Rockland and it takes me that long to drive up and see her, and then I drive back the same day.


How far is that in distance, Jenny? The range of an EV has been one of the big problems in the past with early models of Renault Zoe and Nissan Leaf only managing around 100 miles on a single charge. However, car manufacturers have been pouring money into battery development, partly as a result of the drop in sales of diesel models after the emissions scandal, and so EV ranges have been taking off exponentially. The latest Kia Niro has an official range of 282 miles on a single charge, and this is only going to keep getting bigger as technology develops.

Jenny wrote:
How are we going to generate the electricity to cover the necessary journeys in a green and sustainable way?


This is essentially a separate problem from making electric cars. In order to make sure that EVs produce fewer emissions than ICE cars, you clearly need to invest in renewable energy generation, otherwise you're simply shifting where the emissions are being generated*.

However, the difference between EVs and ICE vehicles is that, as governments buld more and more renewable energy sources, the emissions of an EV will come down over time. By contrast, the level of emissions of an ICE is fixed from the moment it leaves the factory and nothing can be done to change that.


*Not that that's necessarily a bad thing. Lots of reports have claimed that poor air quality in cities is harming people's health. If the bad emissions were created by a power station in the middle of nowhere, it might not save the environment but it would still improve the health of city dwellers markedly.

 
Alexander Howard
1312718.  Wed Feb 06, 2019 10:29 am Reply with quote

Village nuclear - that's the answer. Each town or large village has a small nuclear power plant, like the ones in submarines, with a core light enough to be fuelled off-site and installed and removed by truck. The reactor can be fuelled for life, or taken off-site for new fuel rods to be installed, so it is all sealed in during its operation.

It's less efficient that having a big nuclear power station in the next county and pylons to bring the power, but a smaller core is cheap to refurbish or decommission, and that may be the ultimate saving.

 
PDR
1312720.  Wed Feb 06, 2019 10:43 am Reply with quote

It's not just the rangte; it's the recharge time. The e-Niro is very good, but it still takes nearly an hour to charge from 10% to 80% charge from a 100kW "fast" charger - that gives you another ~200 miles of typical warm-weather, daytime usage (range at night is lower due to the headlight current drain, and range in cold weather reduces partly due to battery efficiency losses and partly becuase people want the heater on!).

Full charges take several hours because (a) the charging rate tapers off as the battery fills, and (b) you can't safely do a full charge at the 100kW "fast" charging rate - this is a feature of the lithium battery technologies. It's also a feature that fast charging reduces the life (number of charge-discharge cycles before the battery degrades) of the battery. I don't know about the E-niro but most of the others like the Nissan Leaf only allow one or two consecutive "fast" charges before you have to do one or more "slow" charges to protect the battery. So you can't just keep fast-charging to do a long journey even if you're happy to be hanging around for an hour.

This is very significant because the battery is an extremely expensive item. The smaller battery of the Nissan Leaf costs a whopping £20k to buy new, and although they are now offering reduced-spec "remanufactured" ones at about 20-25% of this it's still a fair old chunk of pocket change.

It's not currently clear that there are development routes available to make batteries significantly cheaper, or to allow them to charge faster (or more tollerant of fast charging). The nanotechnology electrodes provided the last jump, but at the end of the day the characteristics of lithium phosphate and lithium cobalt chemistries will remain much the same.

The work being done in Formula 1 technology to hot-house hybrid powerplants has yeilded much faster maturing of the technologies to the point where a the intended target of a large car with a 25-30BHP turbo-compounded internal combustion engine equipped with hybrid KERS and 100-150 BHP of electrical power is nearly a practicable option. Such a car should be able to do over 90mpg around town and approaching 180bhp on long cruises. This may be a more practicable medium-term option for general use like Jenny's.

PDR

 
PDR
1312722.  Wed Feb 06, 2019 11:00 am Reply with quote

Alexander Howard wrote:
Village nuclear - that's the answer. Each town or large village has a small nuclear power plant, like the ones in submarines, with a core light enough to be fuelled off-site and installed and removed by truck. The reactor can be fuelled for life, or taken off-site for new fuel rods to be installed, so it is all sealed in during its operation.

It's less efficient that having a big nuclear power station in the next county and pylons to bring the power, but a smaller core is cheap to refurbish or decommission, and that may be the ultimate saving.


A big submarine powerplant develops around 200MWatts*. The general guidance of typical use seems to suggest that a ratio of 650 homes per megawatt is a working estimate for domestic consumption, so one of these plants could supply about 130,000 homes. If you work on an assumption of 10 million homes in the UK that yeilds a need for 80 new nuclear power stations (even if we ignore the commercial/industrial demand and the needs of all these new electric vehicles), each of which needs to be somewhere that has access to a reasonable supply of cooling water. That's not a particularly straightforward business case to satisfy.

PDR

* the associated submarine steam plant isn't rated for continuous use, but you could replace that with a conventional steam plant and turbogeneratorset relatively easily

 
cornixt
1312723.  Wed Feb 06, 2019 11:12 am Reply with quote

For two-car families, having one as short-range commute car is practical. >99% of my car usage is under 50 miles, so while I don't need a full-tank range it is still a bit too close to the mark to get the 80 mile cars.

 
dr.bob
1312726.  Wed Feb 06, 2019 11:44 am Reply with quote

PDR wrote:
It's not just the rangte; it's the recharge time.


That's a fair point, though the two are linked. A longer range means fewer recharging stops for the same journey. If Jenny can make her journey on a single charge, I'm sure she'd have no problem with leaving the car charging for an hour or more while she catches up with her friend.

PDR wrote:
This is very significant because the battery is an extremely expensive item. The smaller battery of the Nissan Leaf costs a whopping £20k to buy new, and although they are now offering reduced-spec "remanufactured" ones at about 20-25% of this it's still a fair old chunk of pocket change.


The cost of the battery isn't the most important factor in EVs. The big thing is the cost of energy storage. If you can produce a battery at the same price as an older one, but which can store much more energy, then the cost per kWh comes down.

The generally accepted wisdom is that EVs will achieve price parity with ICE cars when the cost of energy storage drops to around $100/kWh.

PDR wrote:
It's not currently clear that there are development routes available to make batteries significantly cheaper


Elon Musk seems to disagree with you. At last year's Tesla shareholder meeting, Mr Musk claimed that there were production processes that could help to reduce the price of batteries. Not to mention economies of scale with innovations like their Gigafactory. A recent study estimated that, even if technological innovation reduced significantly, simple economies of scale would help to bring the cost of power storage below $100/kWh by 2020.

 
suze
1312727.  Wed Feb 06, 2019 12:02 pm Reply with quote

Alexander Howard wrote:
Village nuclear - that's the answer.


No one wants a nuclear power plant in her so-called back yard. Even under the current system where nuclear power plants are deliberately built as close as possible to the middle of nowhere, it takes on average thirty years to get a new nuclear power plant proposed, objected to, publicly enquired into, finally approved, and actually built in this country.

Start trying to build them in affluent villages in the Home Counties - which are exactly the sorts of places that would need them under your plan - and you could probably double that. A plan which would really, really, annoy the Campaign to Protect Rural England is not going to come from the Conservative Party.

 
Jenny
1312731.  Wed Feb 06, 2019 12:44 pm Reply with quote

PDR is right - I was conflating two issues in my OP. However, both of those are issues that I was hoping to address in this thread.

I should also have said Bangor, not Rockland - it's about 130 miles from here.

I am sincerely hoping Elon Musk is right about new technology in the pipeline to address the issues of battery storage and recharge time, as well as the energy supply, but I suspect he may be being optimistic.

As a tangent - but an important one - there is the issue of the mining of materials for improved batteries, and their subsequent shortage since some of the materials frequently used are not particularly common.

 
PDR
1312738.  Wed Feb 06, 2019 1:39 pm Reply with quote

dr.bob wrote:
PDR wrote:
It's not just the rangte; it's the recharge time.


That's a fair point, though the two are linked. A longer range means fewer recharging stops for the same journey. If Jenny can make her journey on a single charge, I'm sure she'd have no problem with leaving the car charging for an hour or more while she catches up with her friend.


True enough, although to get a greater range mostly comes from using a bigger battery, and the specific range will then drop due to the greater weight of the larger battery so the gains would not be pro-rata. Also the limiting factor on the recharge time is the capacity of the charger. The standard is currently 100kW for commercial "fast" charging and a bigger battery will take longer to get to the 80% charge point from the same charger.

And you'd need custom home installations even to use the 100kW charger at a house (rather than at a "petrol station" or other car refuelling facility). Domestic supplies in the UK are (as we know) nominal 230volts. A 100kW charger would need to draw over 400A from the grid at household voltage. IIRC the standard household provision is 100A (200A if you have electric-only heating). So Jenny* won't be able to do a 1hr recharge at her friend's house unless her friend also has special power provision for an electric fast-charger. For this to be the norm in every home the national electricity supply system would need to add an additional 400A of capacity per house - a 200-400% increase in capcity that is already close to maxed-out.

Quote:

The cost of the battery isn't the most important factor in EVs. The big thing is the cost of energy storage. If you can produce a battery at the same price as an older one, but which can store much more energy, then the cost per kWh comes down.


I take your point, but if you have a part of your car which costs even £2,000 and has to be replaced every 2-3 years I suspect you' not be happy - I certainly wouldn't!

Quote:

The generally accepted wisdom is that EVs will achieve price parity with ICE cars when the cost of energy storage drops to around $100/kWh.


I would like to see the analysis behind that, because a few quick calcs seem to suggest they are only considering the initial capital cost (ie the cost of building the car with the battery vs an ICE). Current ICE units last at least 100,000 miles and typically 150-200,000 miles, which amounts to 10-20 years in typical domestic usage. Batteries won't even get close to that, and there doesn't seem to be any technologies onm the horizon that might change that. So with the battery being a "consumable" I'm not sure I'm comfortable taking that number without the data.


PDR wrote:
It's not currently clear that there are development routes available to make batteries significantly cheaper


Elon Musk seems to disagree with you. At last year's Tesla shareholder meeting, Mr Musk claimed that there were production processes that could help to reduce the price of batteries. Not to mention economies of scale with innovations like their Gigafactory. A recent study estimated that, even if technological innovation reduced significantly, simple economies of scale would help to bring the cost of power storage below $100/kWh by 2020.[/quote]

Elon Musk says all sorts of things - he hasn't delivered on many of them. But there is major concern at the level of infrastructure expansion we'll need to fund to be able to produce the feedstocks (especially lithium carbonate and cobalt) in the sort of quantities that his claimed manufacturing targets will require. Never mind the environmental issues associated with that level of mining/refinement. And of course we haven't even started on the supplies of selenium cobalt, gadolinium etc for the motors these cars need...

Personally I'd like to see if we can make biohydrogen work before we commit to that sort of thing.

PDR

* yes, I know that Jenny is in the USA, but tthat makes it worse rather than better - a typical US house is supplied with 240v at 50-100A per house, which is then split into two opposing 120v phases by a centre-tapped transformer

 
Alfred E Neuman
1312763.  Thu Feb 07, 2019 4:01 am Reply with quote

My personal perspective on electric cars is in a bit of a state of flux. I like cars, I enjoy driving and I do a lot of mileage. Ignoring for a moment the environmental impact, if I could save the cost of my annual fuel bill, it would make a difference to my lifestyle. But around once a month I drive up to Joíburg and back and do some driving around there, so on any single leg of the trip Iíll be doing up to 700km. When I go on holiday or visit my daughter at university (a few times a year) a single dayís travel is about 800km, often going over 1000km. So Iíd need two cars. An electric for daily use (often around 200-250km a day), and one for road trips. Thatís not going to happen any time soon, because itís just too expensive.

PDR wrote:
True enough, although to get a greater range mostly comes from using a bigger battery, and the specific range will then drop due to the greater weight of the larger battery so the gains would not be pro-rata.

To nit-pick, the extra weight does necessarily lead to a drop in range - at steady speed the mass of the vehicle doesnít affect consumption (all else being equal). Accelerating and climbing hills does though, but then regenerative charting will also be increased, so should offset that to an extent.

On the subject of Elon Musk, letís face it, heís a bit of a bullshit artist.

 

Page 1 of 32
Goto page 1, 2, 3 ... 30, 31, 32  Next

All times are GMT - 5 Hours


Display posts from previous:   

Search Search Forums

Powered by phpBB © 2001, 2002 phpBB Group