The Coming Disruption of Transport

[afransen]

Transit is already largely privatized, going forward. Have you noticed that all of our new transit projects are designed, built and maintained by private consortia?

Your question about AV costs can only be speculated upon. But, if you take for granted that existing Tesla vehicles could fill that role, you could extrapolate from that. A Tesla driven 240km per day with an average trip distance of 7 km and average occupancy of 2 passengers (assuming pooled rides) serves 68 trips per day, or 12.5k trips per year. To match the expected passenger volume of Hurontario LRT would require 2560 vehicles. The base model costs around $52k, so this fleet would cost $133M (equivalent to 0.5km of HuLRT). Of course you would need additional infrastructure, such as a maintenance and storage yard, charging infrastructure etc. Put another $150M for that. Of course, the upfront cost is not that relevant, you are more concerned with per trip cost. Let's be conservative and say that these vehicles only have a 500,000 km useful life, or 5.7 years. That's about $0.10/km for upfront capital cost. Let's double that cost to account for maintenance expense and add charging cost at $0.03/km to get a total of $0.25 / km. Split that over average passengers per vehicle, and each 7 km trip costs $0.88. I mean, you can load it up with a 407-level toll per KM and it's still pretty attractive economics.

However, I see it being more a mix of smaller 2 passenger AVs for individual (non-shared trips) as a premium service, and 12-16 passenger minibuses for pooled rides as the equivalent for 'mass transit' bus, but more attractive service proposition than we see today as in more door-to-door (at least within 100m of origin and destination) and with minimal wait time (5 minute) and higher effective average speed. The opportunity isn't really to eat public transit, though that is likely to happen anyway, but to put a huge dent in private car ownership. Multi-car families go to 1 car, and so on. The average private car travels around 18k km per year. 1 AV can conservatively replace 5 personal cars.

 

[APTA-2048]

Transit is already largely privatized, going forward. Have you noticed that all of our new transit projects are designed, built and maintained by private consortia?

I guess we’ll see how that privatisation works out. But what exactly has been positive? Construction of things like the Crosstown get delayed and go over budget regardless of the DBOM delivery model. As for privatized maintenance and operation, I’m not entirely convinced that’s better either.

Your question about AV costs can only be speculated upon. But, if you take for granted that existing Tesla vehicles could fill that role, you could extrapolate from that. A Tesla driven 240km per day with an average trip distance of 7 km and average occupancy of 2 passengers (assuming pooled rides) serves 68 trips per day, or 12.5k trips per year. To match the expected passenger volume of Hurontario LRT would require 2560 vehicles. The base model costs around $52k, so this fleet would cost $133M (equivalent to 0.5km of HuLRT). Of course you would need additional infrastructure, such as a maintenance and storage yard, charging infrastructure etc. Put another $150M for that. Of course, the upfront cost is not that relevant, you are more concerned with per trip cost. Let's be conservative and say that these vehicles only have a 500,000 km useful life, or 5.7 years. That's about $0.10/km for upfront capital cost. Let's double that cost to account for maintenance expense and add charging cost at $0.03/km to get a total of $0.25 / km. Split that over average passengers per vehicle, and each 7 km trip costs $0.88. I mean, you can load it up with a 407-level toll per KM and it's still pretty attractive economics.

I get it's difficult to compare since it's theoretical. If we go by 6B divided by 32M over 30 years for the Hurontario line to get the $6.25 figure, why not 133M plus 150M divided by 32M over 5.7 years 5.2 times for the driverless vehicles?

However, I see it being more a mix of smaller 2 passenger AVs for individual (non-shared trips) as a premium service, and 12-16 passenger minibuses for pooled rides as the equivalent for 'mass transit' bus, but more attractive service proposition than we see today as in more door-to-door (at least within 100m of origin and destination) and with minimal wait time (5 minute) and higher effective average speed. The opportunity isn't really to eat public transit, though that is likely to happen anyway, but to put a huge dent in private car ownership. Multi-car families go to 1 car, and so on. The average private car travels around 18k km per year. 1 AV can conservatively replace 5 personal cars.

This was still what I was getting at originally. This argument is not the same as the one I was originally replying to. Smaller driverless vehicles are probably going to replace some fixed route service, but it's hard to imagine a complete replacement where there is a lot of ridership. At a certain point you end up replicating a train.

Toronto and Whitby will be trying out a small driverless vehicle to solve the last mile from GO Stations. I'm interested to see how that goes. Getting to and from the GO Station it makes sense. But for the long trip with 1000s of other people from the station into the core, I'm not convinced.

 

[afransen]

For getting downtown, sure. However, that doesn't represent a very large number of trips. Right now transit only really tries to solve maybe 10% of trips in the GTA. For the rest, you have to be desperate to try to use transit. If you're trying to get from Brampton to Ajax, transit will be hopeless to complete with a 16 passenger minibus that can pick up some people in a cluster and drop people off in a cluster.

 

[crs1026]

^I will admit this discussion is opening my eyes, I had never thought of AVs as a transit killer. Some of the arguments are quite compelling.

I can’t see it as particularly desirable, even if the economics prevail.

To use my neighnourhood as an example - we have really good bus service that is very well used, but people have to walk a couple of blocks to reach the larger street where the bus runs. Transit use is probably at or above typical modal share. in consequence, we have walkable back streets with fairly low hourly vehicle counts. Curbside basketball hoops and hockey nets abound, as do speed bumps - the streets are less autocentric than many imagine. If our TTC service is replaced by AV’s, the AV’s will deliver right to the driveway, and the traffic counts in the back streets will rise. (I wonder what will happen when the kids playing road hockey realise that AV’s will politely avoid them, even maybe just sit and wait if they don’t move the net out of the way ;-) )

The road congestion across the GTA is already bad, not just downtown but everywhere - across, say, Highway 7 and Steeles Avenue, up Yonge all the way to Richmond Hill. I cannot imagine adding traffic volume there. While transit volumes fall after evening rush hour, ridership does not die altogether. Instead of roads calming in off hours, adding AV will keep those roads busy into late evening. AVs may be sufficient at 3 AM, but what happens at 8 PM?

The environmental and quality of life impacts of the constant volume of AV traffic, even with quieter electric vehicles, in terms of dust, vibration, and street motion is not where I see us going with “walkable streets”. We are talking about taking back auto lanes to provide better safer bicycle lanes - yes, right out into the burbs. And remember, we are projecting much greater density, not just downtown but across the GTA and likely all Southern Ontario. The impacts are not just “downtown”.

Before we wax poetic about AVs just because they are cool, we need to recognize the downsides and decide how to mitigate those. I’m not naysaying AV’s, but I am certainly raising the con’s for mitigation.

- Paul

 

[2transpo]

You want soulless mega corporations holding our public transit hostage?

Well, if these corporations know transportation then why not? Are you saying that you wouldn't want Elon Musk to have a hand in reshaping transport in Toronto?

How much do you think they could overcharge the government?

Not any more than Metrolinx and their mob contractors do already. You fail to understand how incredibly expensive transit is. AVs on the other hand would run at a profit with fares cheaper than transit, be able to serve all citizens in a city running any route a person requested. With no major setup or construction cost, it is easy to see why AVs are more attractive than transit like LRT.

A future where AV's have decimated our public transit systems is a return to the 70s and 80s and will destroy our urban landscape again. I don't see why you are so enamored with these mega corporations.

What is wrong with the 70s/80s urban landscape? There really isn't anything wrong with it. Sprawl only used to be bad because of pollution and the negative health impacts of longer commutes like stress and lack of sleep. These problems are gone. EVs get rid of the pollution and AVs get rid of the stress and lack of sleep. People should be able to live wherever they want, and if AVs enable that then I am all for it. This whole mentality of cramming people together because that is where the train/bus stops is just plain bad. Let the transport come to the people, and let it take them where they want to go. You certainly have very little route flexibility to go where you want on transit, especially GO rail.

Before we wax poetic about AVs just because they are cool, we need to recognize the downsides and decide how to mitigate those. I’m not naysaying AV’s, but I am certainly raising the con’s for mitigation.

The downsides are very minor compared to the literally world-changing economics of AVs. If we have to accept a few minor inconveniences, that is worth the price of cheap, democratized transport.

 

[crs1026]

What is wrong with the 70s/80s urban landscape? There really isn't anything wrong with it. Sprawl only used to be bad because of pollution and the negative health impacts of longer commutes like stress and lack of sleep. These problems are gone.

Quite the reverse. The loss of farmland in Southern Ontario is a crisis. If there is one problem in Ontario that cries for draconian measures, it's stopping sprawl dead in its tracks. We probably can't reverse recent development.... but if it still has topsoil on it, we should immediately rezone back to farmland. The only way to absorb population growth is to densify where we already have development.

The downsides are very minor compared to the literally world-changing economics of AVs. If we have to accept a few minor inconveniences, that is worth the price of cheap, democratized transport.

Oh, transit isn't "democratized". Right. I see where you are coming from. Let me guess - you just sold your Gamestock options.

- Paul

 

[afransen]

I will admit this discussion is opening my eyes, I had never thought of AVs as a transit killer. Some of the arguments are quite compelling.

I can’t see it as particularly desirable, even if the economics prevail.

Yes. I'm not saying it is desirable--I have described it as the AV hellscape in the past--but it will happen absent policy intervention. I think we should be able to find ways to harness most of the significant benefits of AVs, while minimizing the negative impacts. A congestion charge for downtown might be quite helpful, for instance.

 

[afransen]

What is wrong with the 70s/80s urban landscape? There really isn't anything wrong with it. Sprawl only used to be bad because of pollution and the negative health impacts of longer commutes like stress and lack of sleep. These problems are gone. EVs get rid of the pollution and AVs get rid of the stress and lack of sleep. People should be able to live wherever they want, and if AVs enable that then I am all for it. This whole mentality of cramming people together because that is where the train/bus stops is just plain bad. Let the transport come to the people, and let it take them where they want to go. You certainly have very little route flexibility to go where you want on transit, especially GO rail.

What is wrong with sprawl:

• Consumes huge amounts of environmentally sensitive land, essential for maintaining biodiversity, growing crops, and headwaters for the watershed we rely on for drinking water.
• Low density makes for communities that are hostile to active transportation like walking and cycling
• Low density leads to unsustainably high infrastructure costs. Sprawl is a ponzi scheme, and once it is built out needs to be densified to be economically sustainable on a replacement cycle.

 

[afransen]

I will admit this discussion is opening my eyes, I had never thought of AVs as a transit killer.

And if it's any consolation, it will also be a private car ownership killer, as well as a parking (street, surface lot, garage) and parking minimum killer, which is an unalloyed good thing for cities.

 

[MrGoose]

Well, if these corporations know transportation then why not? Are you saying that you wouldn't want Elon Musk to have a hand in reshaping transport in Toronto?

Yes. I don't want any technology billionaire re-shaping transport in Toronto. Especially not Elon Musk.

Not any more than Metrolinx and their mob contractors do already. You fail to understand how incredibly expensive transit is. AVs on the other hand would run at a profit with fares cheaper than transit, be able to serve all citizens in a city running any route a person requested. With no major setup or construction cost, it is easy to see why AVs are more attractive than transit like LRT.

And we don't want that. Because the cheapness of AVs causes negative externalities that society has to pay such as, congested roads, low rates of walking and urban sprawl. Negative Externalities, ever heard of them??

What is wrong with the 70s/80s urban landscape? There really isn't anything wrong with it. Sprawl only used to be bad because of pollution and the negative health impacts of longer commutes like stress and lack of sleep. These problems are gone. EVs get rid of the pollution and AVs get rid of the stress and lack of sleep. People should be able to live wherever they want, and if AVs enable that then I am all for it. This whole mentality of cramming people together because that is where the train/bus stops is just plain bad. Let the transport come to the people, and let it take them where they want to go. You certainly have very little route flexibility to go where you want on transit, especially GO rail.

As others have said, sprawl is bad because spreading people out like that consumes valuable farmland, parkland, and costs much more in terms of infrastructure. Sprawl is unsustainable, economically and environmentally. And if AV's perpetuate sprawl, then we need to clamp down on them to prevent that.

Route flexibility can be fixed by having higher frequencies and better network coverage. Furthermore, a transition away from car-centered urban design towards an urban landscape that allows for bicycling, walking and transit provides far more route flexibility because now you also have mode flexibility. What you call route flexibility is simply being able to get around by car except now you don't have to drive.

But what if instead of only being able to take the car to the nearest grocery store, you could walk or bike there instead? Encouraging people to walk or bike also has positive externalities because walking and biking is free and encourages a healthy population. A world where everyone takes their AV to do anything is one of obesity and low social interaction.

The downsides are very minor compared to the literally world-changing economics of AVs. If we have to accept a few minor inconveniences, that is worth the price of cheap, democratized transport.

Democratized transport?? How is transit not democratic??? It's a service that is cheap to use that is paid for by everyone.

These "minor inconveniences" you speak of are actually massively negative externalities that include but are not limited to

Low Social Interaction
Unwalkable cities

Ploughing over valuable farmland
Destroying wildlands
Increased Risk of Flooding
Unsustainable infrastructure costs

 

[afransen]

The genie is not going back in the bottle.

 

[crs1026]

The genie is not going back in the bottle.

Very true. The impacts go beyond transport, too. Consider how many jobs are tied to driving, especially first jobs for new Canadians.

- Paul

 

[MrGoose]

The genie is not going back in the bottle.

Banning AV-share services outside of a publicly owned AV transit shuttle in the GTA can help. This isn't like Uber ride-share where there isn't any infrastructure to compete with.

AV-share services would require large garages and fleets of AVs. It is possible to keep them out.

Sprawl repair would also make it much less likely for people to want to drive around.

A possible solution to keeping our cities livable in the future is a temporary (20 years) ban on AV-share services to provide enough time for our policy makers to reduce AV's possible market share.

Another possible solution is a tolling system for AVs based on distance travelled and removing car lanes along our arteries in favor of transit lanes and separated bike lanes.

Essentially, by redesigning our cities around humans and not cars, we can prevent AV's from destroying the urban fabric.

 

[lenaitch]

You are conflating EVs and AVs. They are synergistic, but not the same thing. And all that infrastructure is <$0.02 per km. Tesla model 3 uses ~0.15kwh per km. Electricity in Ontario costs at peak $0.22/kwh. That's $0.03/km at the high end. Of course, transit needs power too.

At some point, we would have to somehow fold in an equivalent to the 'road tax', which is currently about 32% of the pump price for gasoline.

 

[sche]

I've made a summary of what I think of autonomous vehicles. I've tried to be as comprehensive and objective as possible.

First of all, autonomous electric vehicles (which I will refer to as AEVs) in ridesharing fleets are going to disrupt transportation as much as the car itself, perhaps more. This is inevitable.

This is because of just how cheap AEV ridesharing will be for consumers and operators:

• Electric cars have very few moving parts and thus have very little maintenance. There's not much that can go wrong.
• Fuel costs for AEVs will be very low. Electricity costs less than gasoline.
  • If we factor in renewable energy this actually gets even better. Capital costs for solar and wind are getting dramatically cheaper every year (already cheaper than fossil fuels in many cases). Since these energy sources are not consistent year-round, a significant amount of excess energy will be generated on windy and sunny days. Some of this will be stored in batteries, but since solar panels and wind turbines are much cheaper than batteries, it is cheaper to add more solar panels and wind turbines and generate more excess energy than to add large amounts of batteries.
• Autonomous cars will be far, far safer than human drivers, and insurance/accident costs will likewise decrease. AV's can see in all directions at once, don't get drunk, don't get tired, and don't get distracted, eliminating a huge proportion of accidents (human-driven cars are incredibly deadly
• AEVs will also have very low capital costs - all core technology like batteries, cameras, etc. are all decreasing in cost rapidly. Plus, with so few moving parts, they will be able to operate for 1,000,000 mi (1,600,000 km) before needing to be replaced, so the cost of the vehicle is spread out onto so many trips that it becomes almost nothing. With high utilization in ridesharing networks, this can be taken advantage of.
• For ridesharing networks, no driver = biggest cost eliminated. Not much else to say there.

That is pretty much every monetary cost currently associated with cars eliminated or dramatically reduced. This will flip the existing transportation system upside down. All of the above are basic facts, and all it takes is to connect the dots to see the inevitability of this disruption. There is no point denying or trying to avoid this.

This AEV transportation system will have many benefits:

• People no longer own cars (since AEV ridehailing will be cheaper for consumers)
• No more parking needed in urban areas, only drop-off zones.
• No emissions assuming renewable energy grid, or low emissions with current energy grid. No air pollution from cars.
• Subsidized public transit in smaller metros and suburban areas will no longer be needed, freeing up some funds.
• Some increase in road capacity from less lane changing and better drivers, potentially higher vehicle occupancy from ridesharing.

But there will also be enormous downsides, which cannot be ignored:

• Substantial increase in VMT everywhere, due to both new trips and trips displaced from other modes
  • Increased wear and tear on roads will happen
  • A new push for increasing road capacity may happen. New highway construction etc.
• Suburban sprawl. This is the really bad one, if left unchecked. Without the disincentivizing force of driving, sprawl could dramatically increase.
  • This may be somewhat better than current suburban sprawl, with narrower roads, no parking, no emissions, and slightly more bearable commutes. But it will still be terrible:
    • Huge infrastructure costs of building and maintaining huge networks of roads, water mains, fire stations, etc. Low-density sprawl inherently requires far more infrastructure per person than high density neighbourhoods. AEVs don't change this. In many cases the cost of this is so high that when it is time to do major repair or replacement of infrastructure in sprawl areas, the accrued tax revenue from these areas is not enough to pay for it, requiring debt financing or causing infrastructure to be neglected and fall apart.
    • Enormous impacts on farmland (impacting food security), water quality, and local ecosystems.
    • Car dependency in car-centric neighbourhoods make active transportation impractical. The result is 'obesogenic' neighbourhoods, where everyone rides a car everywhere and thus no one ever gets exercise unless they do it intentionally, leading to high obesity rates and associated societal problems.
    • Longer commutes. Even if the commute is more productive than driving, and less stressful, urban sprawl will still create long commutes.

Impacts on urban, high capacity, high ridership public transit is a big question mark for me. AEV ridehailing will likely be more convenient than public transit and probably cheaper under the current system. However, in transit-oriented areas, there is not enough road capacity for all trips to be made on AEVs, and transit is still competitive. Rail lines will still be more effective at moving large numbers of people.

• In the short term, there will be an equilibrium point, just like with cars and transit today. Speed is the most important factor when people choose a mode. Since speed decreases with demand on roads due to traffic (something which might shift a bit with AEVs but does not fundamentally change), demand for road transport increases until road trips are the same speed as transit (or active transportation) trips.
• For the long term, I think there are a range of possible outcomes, and I have no clue what might happen, but here are my worst and best case scenarios:
  • Worst case: Poor AEV accessibility in the central city causes people and jobs to leave, which makes them less vibrant and desirable, creating a negative feedback loop. A new era of urban decline begins, just like what happened in the 1960s-80s.
  • Best case: AEVs make the central city more vibrant and attractive by eliminating parking. Investments in active and public transportation make these areas extremely convenient to get around using those modes, as opposed to AEVs.

Conclusion
AEVs will be incredibly cheap for consumers, and cheaper than public transit today if no changes are made to the current system. However, just like cars, they may not be so cheap from a societal level. Under the current system, AEVs will heavily subsidized, just like cars today - they don't pay for the ROW and road that they use, they don't pay for the infrastructure costs of urban sprawl, they don't pay for the obesity caused by urban sprawl, etc.

But, this is not to say AEVs are unequivocally bad - they do eliminate some of the negative externalities of cars, like time wasted while driving, car accidents, and air pollution. These benefits must be recognized.

However, other negative externalities remain, and one of the worst negative externalities of cars, urban sprawl, may get worse. As such, the solutions to mitigate the problems caused by AEVs will be similar to solutions to mitigate those caused by traditional cars, including:

• A hefty road tax, much higher than current gas taxes, enough to pay for road construction and maintenance, plus some extra.
• Strict zoning policies aimed at intensifying existing areas and preventing suburban/exurban sprawl.
• A first-principles approach to urban planning, instead of a car-centric approach.

The transition to AEVs will be inevitable and unstoppable. It is our job to adapt in order to reap their benefits, while also curbing their problems.