North America

The Double Double


Imagine there existed a Narrow Tram: a streetcar or light rail train that is only about half as wide as the streetcars and LRTs most cities use today.

Narrow trams could fit more easily on narrow streets where there is not otherwise room for a transit-only lane. They might also squeeze into unused edges of existing railway or expressway corridors, where they could avoid red lights. On wider streets, narrow trams could have room for double sets of tracks in each direction: one lane to provide local transit service, the other for express.


For passengers to comfortably navigate such narrow vehicles, each narrow tram could have an unprecedentedly large number of doors – the more doors the better. Each door would open only if passengers indicate they are getting on or off at that precise section of the vehicle. Platforms would tell passengers in advance which sections of each approaching narrow tram are crowded. Fold-up chairs would be located along one side of the vehicle, only one chair per row. Some chairs could be sideways-facing to let people to sit together, others would be front-facing.


Spanish Solution in Stuttgart

In certain cases, narrow trams could perhaps free up enough room for platforms on both their sides – the Spanish Solution – in order to simplify boarding and alighting and allow passengers to transfer very easily between local and express vehicles. Spanish Solution narrow trams might have to have even fewer seats, but this might be manageable, since express trams would be able to travel further in less time than conventional transit, so passengers might be more willing to stand. Plus, with multi-multi-door vehicles and smart-platform systems, you might be able to put chairs close to doors without creating passenger blockages, so the number of chairs might not have to be reduced too much even with the Spanish Solution’s addition of doors along both walls.


Vienna Feldbahn


Narrow tram systems that have separate local and express tracks might also eventually help to facilitate the use of freight trams, at least during times of day when there is not a high demand for passenger transit. With two tracks per direction, a narrow freight tram could linger in one spot to be loaded or unloaded without blocking other trams behind them. If, for example, the process of loading and unloading goods from trucks and trains becomes automated, leading to an increase in multi-modal freight transport or nighttime freight delivery, freight trams might become useful in urban areas because of how quiet, clean, and battery-free they are. Narrow trams might also serve well as freight trams by being able to squeeze into the edges of certain railway or expressway corridors where industrial and freight-transport infrastructure already exists.

Narrow trams might also allow for longer vehicles, at least on their express lines or on grade-separated sections. They could have more room to carve out the wide turns needed by longer vehicles, since their narrow size might allow them to make diagonal cuts to avoid intersections without taking up too much valuable street-corner real estate. Such diagonal cuts would also them to have useful indoor stations. Their longer length could help compensate for narrow trams’ smaller number of passengers per row. Eventually, perhaps, driverless vehicles (at least, in grade-separated corridors) might also allow for viable narrow trams that do not have long lengths.


Glasgow Subway

In some cases, it would not just be the width of the vehicle that would be narrow, but also the diameter of the vehicle. A Narrow-Diameter tram would have both a narrow width and a low ceiling. A ceiling height of seven or eight feet, for instance, would be relatively low, yet would not be so low as to make passengers too uncomfortable, particularly given that the plentiful doors and smart-platform system would help prevent passengers from having to fight their way through a crowded vehicle. The benefits of having a lower ceiling and smaller diameter could be significant. Lower ceilings can make it easier to use underpasses or get in and out of tunnels. Smaller diameters also reduce the required size of tunnels. A tunnel with a 7-foot diameter, for example, would have a volume that is only a quarter as great as that of a tunnel with a 14-foot diameter.

Hypothetically – very, very hypothetically a city could even take an existing subway tunnel and repurpose it to run local and express narrow subway trains in each direction, with the express narrow trains being so narrow that they could able to bypass the local trains at certain points. Even more unrealistically, a city could build a futuristically tubular train with a diameter of only, say, 4.5 feet, in which most passengers would have to sit in order to fit on board, like in a car. A 4.5-foot-diameter tunnel would have a volume only around one-tenth that of a 14-foot one.

Of course, such extreme steps are not needed. There might be benefits to be gained from making vehicles even just a little bit narrower than they are today. Technologies such as digital payment systems that make it easy to board any door on a vehicle, transit apps that make it easy to choose between local or express vehicles, the possibility of smart-platforms that can indicate which sections of approaching vehicles are crowded, and perhaps eventually also the possibility of automation, might all make narrower trams more viable than they have been until now.











North America

Light Rail and Autonomous Vehicles in Toronto

Light rail systems are often a Goldilocks-style compromise between the flexibility of automobiles and the efficiency of trains. The problem is, nobody likes Goldilocks.

If, for instance, Doug Ford is elected premier of Ontario this spring, it is not unlikely that he will cancel the Hamilton, Hurontario, and Sheppard LRTs, leaving only the Eglinton Crosstown and Finch West projects that are already underway. And Toronto’s mayor and city council already voted last year in favour of the suburban Scarborough Subway Extension, over an alternative plan to build that line as an LRT and then use the money saved to help fund an Eglinton East LRT.

On the autonomous vehicles front, meanwhile, a number of significant barriers to entry remain. These include: LIDAR (still very expensive, and still struggles with snow); LIABILITY; the fact that people already own conventional cars; the fact that autonomous cars (even electric ones) still cause traffic and environmental harm; and the risk of autonomous vehicles being used in a terrorist attack (for e.g. if driverless cars are common, a single bombmaker might be able to load numerous vehicles with explosives, and detonate all of them simultaneously at a crowded urban location). And of course there may also be a societal hesistancy to adopt widespread driverless cars.

Because of these barriers, it seems plausible that the partial use of autonomous vehicles will occur before they become fully adopted. Consider, for example, two potential partial usages: autonomous parking lots, and autonomous overnight cargo deliveries. Both of these may not be subject to the barriers listed above:

—LIDAR may not be a challenge for autonomous parking lots, as within a relatively small, mapped area equipped with sensors (the parking lot), cars could drive autonomously without LIDAR. Overnight delivery vehicles might also be able to run without LIDAR, as they could drive at a very slow speed, and stick to running a relatively small number of high-demand routes
—Liability may not be a challenge either, as the parking lot could have no pedestrians or human drivers in it, and its cars could drive at slow speeds. Overnight delivery vehicles could also drive at slow speeds.
—the fact that people already own conventional cars is not a barrier to overnight cargo deliveries, and may not be a barrier to parking lots either. Some companies are even attempting to develop vehicles that can, in effect, tow a conventional car autonomously to and from parking spots
—the fact that autonomous cars still cause traffic and environmental harm may not be a barrier: autonomous parking lots can reduce traffic and pollution if they are located at (for example) train stations, thereby making it more convenient for suburbanites to use transit. And overnight deliveries might cause fewer diesel trucks getting stuck in daytime traffic jams, which create air pollution and other costs
—restricting autonomous vehicles mainly to limited areas like special parking lots, or special times like very late at night, could make it much more difficult for them to be used in a major terror attack (whether a car-bomb/truck-bomb attack or driving a vehicle into pedestrians, involving multiple vehicles simultaneously) as it would then remain suspicious for a driverless truck to be loitering in a crowded urban area
—special autonomous parking lots, and perhaps also overnight autonomous cargo deliveries, are much less likely to be subject to a societal hesitancy towards their adoption

LRTs in particular may benefit from autonomous parking lots and/or autonomous overnight delivery vehicles. Autonomous parking lots may promote transit usage in general, if the parking lots were located at transit stations. But perhaps LRT would benefit from them more than heavy rail would, as the flexibility of LRT relative to heavy rail could allow LRTs to directly access more of these parking lots.

For overnight cargo deliveries, LRTs could be the ideal vehicle to be used autonomously. LRTs are electric and therefore relatively quiet, and being quiet is crucial for overnight usage in cities. Also, electricity prices are cheaper at night than they are in the day (particularly in Ontario, given that the province’s nuclear and wind power cannot shut off at night). And, of course, they are much cleaner than non-electric (or even electric) trucks. In addition, an LRT, unlike heavy rail, could more often travel directly into a building or parking lot to load/unload its cargo.

One main problem that has prevented cargo light rail in the past (outside of a few exceptions, for example in Dresden where a cargo tram has run) has been that trains have less surface friction than wheeled vehicles, so it is difficult for an LRT carrying a heavy amount of cargo to accelerate and decelerate constantly in cities in order to stop for red lights, passenger stops, and — if the LRT is not operating in its own separated lane — cars. At night, however, there are far fewer cars or passenger LRT stops, and green light-red light cycles could be made to run for far longer lengths of time in order to minimize the number of times an LRT has to stop.

With autonomous vehicles, then, LRTs may no longer be only a compromise between heavy rail and autonomobiles, but instead might excel at complementing autonomous parking lots, or being used autonomously to deliver cargo.

What does this mean for Toronto? Well, as mentioned earlier, it is possible that all but the Eglinton Crosstown and Finch West LRT plans may be cancelled as a result of the coming election. The Eglinton and Finch LRTs, as it turns out, have something in common that could be relevant to this discussion: they are next to the city’s two major hydro corridors, the Finch Corridor and the Gatineau Corridor. These corridors could be used as autonomous parking lot systems that are directly accessible to passengers using the LRTs, as well as accessible to passengers using other corridor-adjacent transit stations like Finch Station and Kennedy Station. They would also be accessible to cyclists using the bicycle paths that already exists within these two hydro corridors.

hydro corridor map

Finch Station Parking Lot

If you look at Finch subway station (map above, picture below), you will see that it already has a large parking lot, 1.3 km long and 90 metres wide, within the Finch hydro corridor to both its west and its east. I propose that this lot be extended much longer, to reach north of the Finch West LRT, as an autonomated parking lot corridor. This corridor would mostly remain separate from road traffic and pedestrians, though not entirely separate: it would have to cross north-south streets, and would also have to use bridges on Finch in order to cross topographical barriers like G Ross Lord Park. But that would still be much less of a challenge than a widespread adoption of autonomous vehicles. The Finch corridor is about 210 metres north of Finch in most places, and in some places (such as west of Jane, or west of Bathurst, or between Dufferin and Keele) it widens to connect to Finch Avenue directly.

finch station parking lot.png

Finch Subway Station (Yonge and Finch)

The Gatineau Corridor, meanwhile, intersects with the Eglinton Crosstown just west of Victoria Park, and also (via the narrower Scarborough RT corridor; see bottom image below) at the Crosstown’s terminus station, Kennedy Station (which is also a station on the Bloor-Danforth subway, Scarborough RT, and Stoufville GO train). If the Eginton East LRT extension to the Crosstown is built, its terminus would also be by the Gatineau corridor, at U of T Scarborough campus.

Where Eglinton Crosstown and East LRTs meet hydro corridor.png

Above: 3 Locations Where Gatineau Corridor Meets Eglinton (or Eglinton East) LRT; Below: Kennedy Station

Kennedy Station .png

The corridor could be relatively quiet, since the cars parking in it could travel slowly. It would not be an eyesore; or at least, not more of an eyesore than the hydro towers are themselves. It would also, ideally, be “parking lot neutral”; in other words, by creating more parking on the hydro corridor, it would allow you to convert some existing parking lots elsewhere into buildings/parks/etc. It would promote an increase in transit ridership. And the corridor could also be used, seasonally, as a “bicycling highway” that would be usefully located next to the autonomous parking lot. This could be acheived by simply having a portion of the hydro corridor’s lanes be designated for cycling instead of parking during the warmer months of the year. This could be a transit option that both the suburban, car-driving Ford Nation and the latte-drinking downtown bicycle-lovers could enjoy.

North America

The Double Double LRT – Make Hamilton Great Again

In 1870, Hamilton had a population half as large as Toronto’s. If it were to have kept that up, Hamilton would today be the 3rd largest Canadian city. The Bulldogs might be competing for the Stanley Cup right now.

Toronto-Hamilton-Buffalo Populations

But, Hamilton has been held back by two processes. One was the move from water transport to land transport. Hamilton’s main advantage was its port, which is still by far the busiest among Canadian ports in the St Lawrence/Great Lakes network. But in the modern era dominated by land transportation, Hamilton has been limited when compared to Toronto, as it is somewhat blocked in by the Hamilton Harbour or Niagara Escarpment (or, more distantly, by Lake Erie) on all sides.

Hamilton port

The other was the move move away from manufacturing, and towards service-oriented metropolises. This too has held back Hamilton relative to Toronto, as Hamilton’s port (and its position between Lake Ontario and Erie, near the Welland Canal) had made it a manufacturing leader.

But things may be changing again:

  1. Many services that cities like Toronto specialize in may face outsourcing or automation
  2. As industry continues to automate, it may allow some “re-shoring” to high-wage countries like Canada. (And, if services are automated or outsourced, wages may fall in countries like Canada, relative to other countries)
  3. If fuel prices (or pollution concerns) increase, it may lead water transportation (for cargo) or re-shoring manufacturing to become more important
  4. If machines remain unable to compete with the dexterity of human hands, manufacturing may need to remain located in cities, with human workers operating alongside machines
  5. E-commerce and automated warehousing may free up some urban commercial real estate for industry

In other words, it is becoming increasingly possible to imagine that mid-sized port cities like Hamilton, and large service cities like Toronto, will both reindustrialize to a certain degree going forward.

Indeed, cities in Ontario may be particularly well-suited for (partially) automated industry. Ontario has the highest disparity between surplus overnight power (generated by nuclear plants and wind farms, which cannot turn off at night) and relatively expensive daytime power. This disparity automated factories might be able to take advantage of, as machines can run overnight.

What follows is a joint transportation plan for both Hamilton and Toronto, aimed at transporting people during the day and transporting cargo during the day and night, while maximizing fuel-efficiency in both cases.

The Double-Double

Cargo is heavy, and so requires a lot of energy to transport. Luckily, as mentioned above, Ontario has a lot of surplus nighttime electricity (especially on windy nights). Overnight is also the best time to transport cargo, because there is no traffic on the roads and because green light-red light cycles can last far longer than they can during the day. This is important for transporting cargo in a fuel-efficient manner, as the weight of the cargo means that acceleration and deceleration requires a lot of fuel. Being able to drive at a constant speed without having to constantly stop and start because of traffic jams or red lights is a big benefit. Today overnight transport is held back because it is expensive to pay workers to drive vehicles, or to load and unload vehicles, overnight. But if machines are doing the driving and loading/unloading, that could change.

Electric vehicles are ideal for overnight transportation, not only because of Ontario’s surplus overnight electricity, but also because electric vehicles are quiet — and being quiet is obviously very important overnight. But electric trucks are simply not efficient enough. Light rail would be better.

Overnight transportation can help solve one of the main challenge faced by light rail/streetcars: namely, the fact that are less efficient at accelerating and decelerating than wheeled vehicles are (because there is less surface friction for rail than road), so having to stop and start constantly because of traffic and red lights makes light rail/streetcars much less efficient than they would be at night when there is no traffic and far longer green lights. This is particularly true of longer, heavier LRTs that would otherwise be more efficient, for example those that will be used by the Eglinton Crosstown (which will be able to handle three of the new long Toronto LRT vehicles strung together to create a really long LRT train). Longer LRTs are also slower at turning; overnight, however, they could turn slowly without causing a traffic jam of cars waiting behind them.

Cargo LRTs could also capitalize on one of light rail’s main advantages over heavy rail: flexibility. LRT tracks can branch off of the main route to travel directly into an industrial building, in order to be directly loaded or unloaded. a

But what about the daytime? How do you maximize light rail efficiency when a vehicle in the daytime would have to stop constantly for red lights, or stop to drop off/pick up passengers, or stop (if it is forced to share a lane with cars) because of traffic jams? Toronto’s answer to this question has been the Eglinton Crosstown tunnel. But Hamilton obviously cannot afford a subway like the Crosstown. It’s possible that even Hamilton’s surface LRT plan will soon be cancelled by Doug Ford.

In theory, there are two ways to maximize LRT fuel-efficiency. One is an express method: allow an LRT to run quickly in its own lane (not shared with cars), and have as few red lights and pickup/drop-off spots for passengers as possible. This is obviously difficult to do in a downtown setting in the daytime, without a Crosstown-like tunnel. Moreover, unless you have two lanes in each direction, to allow an express LRT to overtake a non-express LRT, the lack of pickup/dropoff spots on the express LRT would mean that the LRT might be less accessible and cost-efficient.

The other method would be to have the LRT travel as slowly as possible, in order to reduce the amount of fuel need to constantly accelerate and decelerate, and allow it to share its lane with cars. But here too, while fuel-efficiency might be maximized, travelling at a slow speed might also reduce cost-efficiency, since it would carry many fewer passengers per hour than a faster service.

The solution to these daytime challenges (if there is one), would seem to me to be to do the following: have two parallel LRT lines on the same street, one a Really Fast Lane and the other a Really Slow Lane.

The Really Fast Lane would run express in the middle lanes of the street, and would not have to share its lane with cars. It would make all of its passenger pick-up/drop-off stops in the Really Slow Line, so that trains in the fast lane would be able to minimize their stops by running express routes. (Overnight, having two lanes would also allow for one of the lanes to be used for loading/unloading cargo, without causing backups).

The Really Slow Lane, on the other hand, would overcome its slowness problem by doing the following: share its lane with cyclists, with cars, and allow cars using the Really Slow Lane to drive autonomously. By driving at really slow speeds, you can allow for autonomous driving without having to pay for LIDAR (and LIDAR has challenges with snow anyway) and with a reduced chance that fatal accidents involving autonomously-driving vehicles will occur, and you can allow cyclists and cars to share the same lane safely and comfortably. As such, you can potentially make a slow lane not only more fuel-efficient than a normal lane, but also cost-efficient.

So, that’s the basic idea behind the Hamilton Double-Double: one Really Fast Lane, for express LRTs carrying passengers or cargo, and one Really Slow Lane, shared by LRTs travelling slowly, fast-lane LRTs making stops to pickup/droppoff passengers or (at night) cargo, cyclists, cars, and autonomous (/advanced cruise control) cars.

For downtown Toronto, however, where the density of red lights and passenger stops is so high that even a Double-Double would not be able to maximize fuel-efficiency (let alone cost-efficiency), since the fast lane would still have to stop too much, while the slow line would be too crowded by traffic, a somewhat different plan could work: a plan more similar to the Crosstown.

Okay, this idea is pretty crazy, I admit, but here it goes:

The problem with subway tunnels in downtown Toronto has been one of expense: they would (unlike the Crosstown) have to tunnel through bedrock, and avoid a lot of underground utility lines and nearby buildings that exist. But what if, instead of having one subway lane in each direction, you only have one lane? What if, on Queen Street, you have 2 surface LRT lanes (one really slow lane and one really fast lane), and one underground lane (travelling in the opposite direction as the surface lanes)? Not only would you have half as much tunnelling as a conventional subway, and more easily avoid underground utilities and buildings, but you would also then have room underground to create turning lanes, so that the underground LRT could branch off to go directly into nearby buildings for the loading/unloading of cargo overnight. So long as the two surface lanes were fast enough to prevent the tunnel lane from being underutilized, maybe this could work….?

Finally, a last crazy idea: connect the LRT systems of Hamilton and Toronto, by having LRTs that can drive directly on and off ships travelling between the Hamilton port and Toronto Port Lands.