Integrating Autonomous Vehicles Into Complete Streets
Austin just released a new set of complete street designs and they are look great. Complete streets are streets that are designed for a wide variety of uses and transportation options. American cities and small towns are embracing complete streets at a breakneck pace and our cities are becoming better places because of it. The only thing that is moving as quickly in transportation evolution is autonomous vehicles. As a thought experiment I combined the two.
Note, this is a very generous right of way (available width for the street), but you can shave a lot of width off of it by reducing median width, vehicle lanes could be skinnier even the bike lanes could easily go down a few feet. Up to 100 feet wide you could have all of these elements easily.
In order to show you the difference, here is that same street in a traditional complete street layout.
There are two major differences I would like to highlight as best practices for autonomous vehicles going forward.
Separation of autonomous vs traditional cars
In the autonomous street layout above both types of traffic are separated. This is important because autonomous cars become incredibly more space efficient than traditional cars when they are the only vehicles driving together. The IEEE estimates that lanes can support 3.7 times as many autonomous vehicles per lane. In order to fully benefit from the autonomous vehicles they will need to be separated. Of course this will not be possible on many roads but it need not be initially. As long as there are major thoroughfares that provide separate facilities for autonomous vehicles, travel times could be greatly reduced and then the cars can mix in with traditional traffic for the last few blocks. Similarly we could time lights (or have smart lights) that provide minimal delay. Even if these complete streets have speeds limited to 35 mph, if the cars can maintain 30 mph it will be a huge improvement in commute time in our congested urban areas.
A shared passing lane for autonomous vehicles
Autonomous vehicles are going to be as close to traditional cars as the automobile was to the horse drawn buggy. Therefore many things that are impossible for human drivers should prove second nature for autonomous vehicles. In this example we have a shared passing lane on a high traffic street. You would never ever allow this for drivers in cities. You do see turn lanes, but this is not a turn lane, this is for delays in the primary lane of travel. The autonomous vehicles could effortlessly move around delays. With efficient car to car communication they could take turns in the most efficient manner if there were delays on both sides. Additionally, this could allow riders to be picked up and dropped off along the medians in this diagram without slowing traffic at all. By placing pick-up and drop-off locations staggered and restricting them to those areas ( these could be extremely frequent) autonomous vehicles would never have to stop due to traffic.
What about cross traffic and pedestrians?
These will still be the major limiting factor in achieving maximum efficiency. Though the delay will be temporary. Once an autonomous vehicle enters the center lanes they will become part of a traffic wave. Much like the timed bike signals in San Francisco and Europe, autonomous vehicles would be part of these timed waves that would never stop until they leave the street. Similarly, during the breaks in the wave pedestrian signals would trigger allowing individuals to pass the street safely. In the event of a pedestrian trying to run across traffic, the autonomous vehicles would handle that better than humans.
Though this is merely a rough outline of what a complete street could look like with autonomous vehicle integrated into it, we need to start planning for it. The future of the streets that surround us depend on it.