The Coming Disruption of Transport

[afransen]

In theory, with the trucks being automated, you could restrict long-haul passage to after 7 pm and before 7 am in major centres.

The irony is that many parts of the city actually have a noise curfew that preclude deliveries during those times!

 

[lenaitch]

A further 12,000 vehicles added to 50,000 is a tad shy of 25% - hardly a "modest Increase". Converting rail ROW to a dedicated AV ROW (privately owned of course because why would the state own it) - I wonder how that impacts the economics.

 

[nfitz]

The irony is that many parts of the city actually have a noise curfew that preclude deliveries during those times!

In residential and downtown urban areas - sure. But they aren't the origins or destination of the kind of long-haul operations that will initially benefit from this.

While we might only be a decade or so from seeing automated vehicles leaving from terminals on/near highway ... I really don't see even cars - let alone more dangerous vehicles, operating without humans inside in dense urban areas and downtown streets until until the 2040s, at a minimum.

Though it would be fun to have things that way, and never have to worry about crossing a street mid-block again!

 

[afransen]

In residential and downtown urban areas - sure. But they aren't the origins or destination of the kind of long-haul operations that will initially benefit from this.

While we might only be a decade or so from seeing automated vehicles leaving from terminals on/near highway ... I really don't see even cars - let alone more dangerous vehicles, operating without humans inside in dense urban areas and downtown streets until until the 2040s, at a minimum.

Though it would be fun to have things that way, and never have to worry about crossing a street mid-block again!

To be fair, you have to worry about crossing mid-block until they ban manual vehicle operation.

It seems like a bold bet that it will take 20 years for autonomous cars to arrive given Google has one in revenue service today. For context, 20 years ago, AOL dialup was how most people got on the internet and the first GPUs was released. And technological advancement is accelerating... I would grant maybe it takes 20 years for AVs to completely conquer the market (>90% share of VMTs). 20 years is a long time, at least it tech.

 

[nfitz]

To be fair, you have to worry about crossing mid-block until they ban manual vehicle operation.

Ah, but currently I only cross mid-block when there's no traffic coming. If it was 100% automated, I'll be able to cross a 2-lane street almost anytime - and even start crossing the first lane, while waiting for the second lane to get a gap, because the car would then have to wait for me. Something to entertain me in my 90s ...

It seems like a bold bet that it will take 20 years for autonomous cars to arrive given Google has one in revenue service today. For context, 20 years ago, AOL dialup was how most people got on the internet and the first GPUs was released. And technological advancement is accelerating... I would grant maybe it takes 20 years for AVs to completely conquer the market (>90% share of VMTs). 20 years is a long time, at least it tech.

Seems closer to 30 years than 20 since AOL got access to the Internet as far as I recall ... a big drop in the signal:noise ratio if I recall (about 1994 if I were to guess).

Less than 19 to be precise ... but yeah, I'll be surprised if there are fully automated cars running regularly with absolutely no one on board that can access any driveable street in downtown Toronto before 2040. Lesser automation that starts beeping and comes to a stop with the need for a human to take over is more likely before then.

 

[crs1026]

The 401 carries over 50,000 trucks per day at its busiest section. If we were generous and and say that between Windsor and Montreal there were 40 trains and each of those trains had 100 cars on them and each car carried and average of 3 trucks then we have the equivalant of 12,000 trucks per day on the rails. That is a modest increase and one that could be handled. Some rail infrastructure like right of ways could even be repurposed for trucks.

You are way off in your numbers. Double stack trains run to 15,000 feet - at 50 feet per container, that’s 600 containers per train. Now figure out how many containers in a 15,000 foot bulk oil train. (and consider the added transload cost of using 3 smaller tank containers in the place of one large capacity railcar).

Take a look at the highway congestion between Kitchener and Oshawa. The only way one could add capacity is through a rigid scheduling system that matches shipments to throughput. That adds time penalties, and could create a need for layover/staging yards. By the way, in the wee hours when road maintenance is going on, it’s not uncommon to find a mile-long creep zone on the 401 as three lanes of traffic merge into one open lane.

Now apply that to the TCH which isn’t 4 lanes end to end, and parallels even busier railway lines.

The video posted above is impressive - certainly the driving prowess of that truck technology is coming along well. But, the application is fundamentally “open road”. I didn’t see how it changed throughput especially in those heavier volumes where you have one or two solid lanes of trucks.

It is simply wrongheaded to add to the congestion on our roads, and it is doubly wrongheaded to try to enlarge highway capacity especially in Southern Ontario.

- Paul

 

[afransen]

Ah, but currently I only cross mid-block when there's no traffic coming. If it was 100% automated, I'll be able to cross a 2-lane street almost anytime - and even start crossing the first lane, while waiting for the second lane to get a gap, because the car would then have to wait for me. Something to entertain me in my 90s ...


Seems closer to 30 years than 20 since AOL got access to the Internet as far as I recall ... a big drop in the signal:noise ratio if I recall (about 1994 if I were to guess).

Less than 19 to be precise ... but yeah, I'll be surprised if there are fully automated cars running regularly with absolutely no one on board that can access any driveable street in downtown Toronto before 2040. Lesser automation that starts beeping and comes to a stop with the need for a human to take over is more likely before then.

Frankly, I'm baffled why you believe this. The interventions, if any, are handled by a remote human operator, not the passenger in the vehicle.

Very long, but if you're curious several mostly contiguous autonomous rides:

 

[crs1026]

Very long, but if you're curious several mostly contiguous autonomous rides:

Thanks for this.

I’m impressed, but I would say I only saw about 40% of the functionality that I would hope for. The test was illustrative but not a ready for prime time version. Perfect weather and road conditions; new, well marked roadways laid out with fairly modern design standards; minimum of unique roadways; no emergency vehicles; few if any pedestrians bicycles motorcycles; fairly sedate driving environment; no malfunctioning traffic lights; virtually zero sudden decisions. Seemed to drive aggressively through stale yellows, couldn’t tell whether it spotted a road work zone with temporary reduced speed signage.

The “pull in to pick up” functionality was fairly primitive. Maybe the roadway can tolerate a single vehicle waiting for its passenger to find it (somewhere it said the vehicle would wait 6 mins before aborting a wait sequence) but if there were a half dozen all waiting, would that clog things up? There may be new roadway designs needed.

Having said that, the technology is coming along nicely, but I would still say years away.

- Paul

 

[Admiral Beez]

Say you ran a trans-Atlantic ocean line in the late 1940s. Would it not be a wise move to invest in and look into operating aircraft in the future in order to phase out your trans-Atlantic ship crossings?

My very first job in 1996 was with Canadian Pacific Limited, when they owned ships (both freight and passenger), an airline (both cargo and passenger), ground transportation (both rail and truck), and hotels. Within a few short years most of this was sold off, including my unit at CP Express.

 

[afransen]

Thanks for this.

I’m impressed, but I would say I only saw about 40% of the functionality that I would hope for. The test was illustrative but not a ready for prime time version. Perfect weather and road conditions; new, well marked roadways laid out with fairly modern design standards; minimum of unique roadways; no emergency vehicles; few if any pedestrians bicycles motorcycles; fairly sedate driving environment; no malfunctioning traffic lights; virtually zero sudden decisions. Seemed to drive aggressively through stale yellows, couldn’t tell whether it spotted a road work zone with temporary reduced speed signage.

The “pull in to pick up” functionality was fairly primitive. Maybe the roadway can tolerate a single vehicle waiting for its passenger to find it (somewhere it said the vehicle would wait 6 mins before aborting a wait sequence) but if there were a half dozen all waiting, would that clog things up? There may be new roadway designs needed.

Having said that, the technology is coming along nicely, but I would still say years away.

- Paul

The point is, it doesn't seem to jive that it will take >20 years to go from this to handling more challenging environments.

I'm just trying to help people calibrate their expectations. It seems like a lot of people are going to be shocked by the change.

 

[crs1026]

The point is, it doesn't seem to jive that it will take >20 years to go from this to handling more challenging environments.

I'm just trying to help people calibrate their expectations. It seems like a lot of people are going to be shocked by the change.

Yes, it seems that people are polarised. Some talk as if the technology will be on our streets in the morning, others are skeptical that it will ever get here.

Increasingly I'm somewhere in the middle.... as I see more of it, I'm taking it far more seriously. On the other hand, I'm struck by how it is still at the test bed stage, is short of being proven as an all-weather, all-roads proposition....and lacks integration beyond the machine itself.

A good example is hailing stands and parking lots....it's quite possible that as the technology is rolled out, different modes of "cab ranks" and "kiss and ride" platforms will be needed. The local shopping mall may need to rework its roadways - parking lots may get smaller, but more space for waiting islands and integration between physical points and the cars may be needed. Imagine a busy day at Yorkdale with loads of shoppers coming out simultaneiously looking for *their* waiting vehicle. Not impossible, and many creative ideas will emerge.... but what if Mall Owner X wants things done one way and Mall Owner Y wants them done another? These things don't shake down overnight, and lots of concrete may need to be poured.

- Paul

 

[nfitz]

Frankly, I'm baffled why you believe this. The interventions, if any, are handled by a remote human operator, not the passenger in the vehicle.

I'm baffled at what you think I believe here.

To be clear - I don't believe that such a situation would be allowed in downtown Torotno until at least 2040. Of course a passenger wouldn't be dealing with it. As I don't think they'd be permitting it, then there'd always be someone in the car behind, or at least near, the wheel.

I think what you are suggesting is that I thought they might allow the use of remote operators before 2040? I don't.

 

[afransen]

I'm baffled at what you think I believe here.

To be clear - I don't believe that such a situation would be allowed in downtown Torotno until at least 2040. Of course a passenger wouldn't be dealing with it. As I don't think they'd be permitting it, then there'd always be someone in the car behind, or at least near, the wheel.

I think what you are suggesting is that I thought they might allow the use of remote operators before 2040? I don't.

Alright. Well, all I can say is hang onto your hat!

 

[2transpo]

You are way off in your numbers. Double stack trains run to 15,000 feet - at 50 feet per container, that’s 600 containers per train. Now figure out how many containers in a 15,000 foot bulk oil train. (and consider the added transload cost of using 3 smaller tank containers in the place of one large capacity railcar).

I'm taking averages. How many trains are actually 15,000 ft in the Montreal to Windsor corridor? Most are usually 5000-9000 ft. So you take the averages and that is what I got, and I was being generous with my numbers.

Take a look at the highway congestion between Kitchener and Oshawa. The only way one could add capacity is through a rigid scheduling system that matches shipments to throughput. That adds time penalties, and could create a need for layover/staging yards. By the way, in the wee hours when road maintenance is going on, it’s not uncommon to find a mile-long creep zone on the 401 as three lanes of traffic merge into one open lane.

Alternate routes exist. The 403 and QEW are available. Plus Highway 413 will ease congestion. Don't forget the capacity and throughput autonomous vehicles will unlock.

The video posted above is impressive - certainly the driving prowess of that truck technology is coming along well. But, the application is fundamentally “open road”.

Yes, and by 2024, maybe even earlier, many trucks on the highway will not have a driver. I suspect a child born in a year or so will not remember a time when trucks had drivers.

It is simply wrongheaded to add to the congestion on our roads, and it is doubly wrongheaded to try to enlarge highway capacity especially in Southern Ontario.

Why? They are up to the task. As Ark Invest showed in their analysis, even with having autonomous trucks pay the full cost of road useage, they are still a better value proposition than rail.

The important thing is that when the trucks pay the full costs, the negative externalities go away and there is no reason to oppose increased truck traffic.

I will reiterate this. Shippers like trucks. They are faster and more flexible and if they can do the job at the same cost as rail or lower, then we shouldn't try to stop it. "But I like trains" is not an arguement against autonomous trucks.

CN is being very forward thinking in this regard. 30 years ago, Algoma Central got out of the rail business and now is primarily a lake freighter line. I could see CN doing the same but moving to trucking.

This would be like if the Cunard Line became an airline, or the Pony Express started doing telegraph. By investing in TuSimple, CN is getting ahead of the disruption and preserving their business and place in logistics and for that I applaud them.

 

[crs1026]

I will reiterate this. Shippers like trucks. They are faster and more flexible and if they can do the job at the same cost as rail or lower, then we shouldn't try to stop it. "But I like trains" is not an arguement against autonomous trucks.

If the argument were founded in liking trains, we would have ceased it several pages back.

What is your prognosis for Seaway shipping? Presumably if the grain is all brought to Thunder Bay by AV truck, instead of train, it will continue on to Sorel rather than transloading to boat?

- Paul