Another look at PLP/Architecture's CarTube

© PLP/Architecture

When we first covered the CarTube, a proposal by PLP/Architecture, a Totally tubular solution proposed for putting cars underground, I was somewhat dismissive, thinking it expensive, impractical, and just another way for rich people to sit in a private little bubble, this time under London. I still am dismissive, given that according to Andrew Gilligan, previously cycle Czar for Boris Johnson, only about 7 percent of Londoners actually travel by car or taxi in the centre of town, making this a very elaborate and expensive 7 percent solution. However, having their brochure in hand, I feel it is deserving of a revisit because it raises some interesting ideas.

platooning© PLP/Architecture
A critical misunderstanding in the original post was my statement that these cars were traveling on a belt of some kind, as I misinterpreted the Guardian article that mentioned that the cars are “locked into position on constantly moving tracks.” They are not; the cars are in fact autonomous electric vehicles (AVs) that are “‘virtual trains’ where vehicles are digitally locked into ‘trains’ with minimal separation. This technology already exists and is being tested by all major car manufacturers.” So if you have the AVs and they are capable of platooning already, then much of the technological heavy lifting is already in place.

netowrk© PLP/Architecture

The real innovation is the proposed network, the grid of pipes that the cars go through. It is a thick mat with a primary grid where the traffic never slows down;

…it constantly runs at a steady speed. All stations are on spurs and are bypassed by the primary grid. This is in stark contrast to the main limiting factor of most modern mass transport systems where the train stops at every station thus limiting its capacity.

tunneling© PLP/Architecture

That’s a lot of pipes, but they are not very big in diameter and the architects have a plan for that too:

Tunnelling equipment will be wholly automated with no need for local operators. Novel robotic techniques are being developed so that tunnelling can proceed continuously with a backload system for removal of spoil and supply of construction materials. 3D printing will be developed to allow shallow curvature tunnels as well as tunnel junctions.

Some critics have complained that it’s pretty crowded underground in London, with centuries of building and plumbing and tube construction. However a search on the geology of London shows that the city sits on a bed of clay, so they can probably just go under it all.

London UndergroundBritish Geological Survey/Public Domain

London Clay is an easy material to tunnel through. This helped London to build its underground rail network quickly in the late 19th century. Another advantage of London Clay is that, clay being impermeable, groundwater does not penetrate into the tunnels.

CarTube - Simulation from PLP Architecture on Vimeo.

In the end, it is still a silly idea for a seven percent solution, with a bunch of people sitting alone in their cars watching videos, but cannot be so easily dismissed as I originally did, because there are some critical insights.

-As autonomous vehicles become common, it is likely that we will have to rethink our road systems and give them a dedicated right of way, much like we do with trains. I have been concerned that a proliferation of AVs would lead to significant fencing and other barriers for pedestrians, or even separating pedestrians with grade separations. like those proposed in Futurama by Norman Bel Geddes in 1939. Putting them in tunnels is certainly a better solution. As PLP/ Architecture's Lars Hesselgren told the Guardian,

The real issue is control. Trains and tubes have controls; cars don’t. Autonomous cars do, but they don’t work well with pedestrians, cyclists and other unexpected elements. The only way you can have a high-capacity car network in a city is to have a dedicated track.

-if you are going to bring this new mode of transportation into the city, you are going to need a lot of pipes or it just won’t work. Much like a good bike lane system, you need a minimum grid. You need a critical mass of routes to get people where they want to go efficiently and safely.

If nothing else, the proposal demonstrates that adapting our cities to handle self-driving cars is likely to be hugely complicated and expensive, and that perhaps Mayor Khan’s investment in bikes is a smarter move.

Another look at PLP/Architecture's CarTube
It might not be as wacky as we first thought.

Related Content on