Traffic Engineer Fantasies

Traffic engineers love connected vehicles because they are so much fun to model. It's like a structural engineering being granted a new super-material--you can do things that used to be impossible. 

Traffic engineering is an engineering discipline and accordingly is empirically grounded in physics. Cars take up space, they have mass, there are limits to how fast they can accelerate, how fast they can stop. Some of those limits are physical, some are averages of human capability. After much empirical work, much is known about those limits, and it's possible to combine what is known to make models and operationalize those models as mechanical simulation models. 

But those same models show limitations -- you can't get more than 2000 cars per lane per hour, because the faster people go, the less time they have to react when the car in front of them brakes, and so the more space they leave. There is an equilibrium, and the optimum speed for maximum flow volume is remarkably low. 

Connected and Autonomous Vehicles (CAV) let traffic engineers throw all of that empirical work out the window and ask "If we automated all the cars, how good could we get?" And the answer, unsurprisingly, is a lot. If we automated all our vehicles, it would save a huge amount of time. 

This has led to a lot of worthless pie-in-the-sky papers being published, as traffic engineers explore the fantasy of complete automation. We know this particular form of naval-gazing is worthless, because transit did the same thing decades ago, of how great things would be once a huge region-spanning network was built. (They weren't). 

Same fallacy: something that works once everything is done is a dead letter, because the scale of the required change is so great that it has to be incremental and evolutionary. Which in turn requires each intermediate stage to be immediately viable and sustainable, as there is never any certainty how long that stage may last. (We can't move the vehicle fleet from gas to electric for the same reason).

The useful papers have been about those incremental stages, rather than the fanciful end-stage. And those papers strongly suggest that the gains from the intermediary state are pretty limited. Indeed, so limited they may not justify the investment. 

So CAV are a bit of a dead letter, for the same reason that electric vehicles have taken so long to spread--infrastructure costs. No one is willing to take the risk without the infrastructure, and no one will invest in the infrastructure without the market basis.

Still fun to run models about, though, and to dream about. 

Bike like someone is trying to kill you

They aren't. (Probably). But pretending someone is will probably save your life. Because it forces you to be cautious and aware. When you get hit, it's going to be an accident, but you are still going to hurt, maimed, or killed. And the driver will say "Oh! I'm so sorry, I didn't see you!" I've been that driver and I've been the "I-need-to-sit-down-for-a-moment---I-almost-just-died" cyclist. 

It's not like the drivers are trying, or that they don't care. But it is a hard reality that our transportation system barely considers the safety of non-automobile users, and discounts their existence as a matter of practice, to shave fractions of sections at traffic signals. So vulnerable users are constantly at risk. It's like swimming with whales, great beast that might accidentally crush you in a moment of inattention. So, cycling in that context requires a certain amount of devil-may-care vitality, that only young men can fully enjoy.  So, when they are the only cyclists on the road, it's not by accident--it's selective pressure. 

Concordantly, the inevitable effect of being willing to bike in a context where every car might kill you is that you become a very aggressive cyclist. It's a natural consequence of living embedded in traffic violence. As a cyclist, the suggestion that traffic laws exist is absurd. Enforcement is, at best, nominal. Traffic violations are endemic. Everyone speeds, and people regularly run red lights, make illegal turns, ignores signs, violates crosswalks, fails to yield. It's a jungle out there. 

Certainly, there are customs, there are norms, there are signs and signals which are generally followed. But there is absolutely zero certainty of that happening, and the violations will happen suddenly, without notice. If we were sitting in a therapist's office, the therapist would nod now and say "Yes, is the very essence of abusive". 

If you do the legal thing, cross at the crosswalk, is there any safety in assuming that someone will come to a complete stop at the red light, before making their right-turn-on-red? Of course not. So you learn to bike in ways which are safe, in the same way people learn to hold a knife safely--make it impossible to get hurt and forget anything else. So you bike on the wrong side of the street (so no one can hit you from behind). You bike on the sidewalk (so you are protected by the curb). You cross the street before the crosswalk timer starts, before the light turns green. You wait for the light ahead of the stop line. 

Don't take anyone who talks about cyclists violating laws seriously. They are upset about cyclists behaving with the same level of privilege and immunity that drivers do, and that upsets their sense of the natural hierarchy of things.