What happens when a new form of mobility is introduced?

No one knows when, how or even whether autonomous vehicles will arrive as a significant real world transportation option.  This lack of certainty creates space for technologies to be developed and tested, and time (at least a little) to think through what might be coming.

Before summarizing a few visions of an autonomous car future, I’d like to jump back a century for a quick review of what happened when human-driven motor vehicles were first introduced in significant numbers.

Los Angeles provides a useful case study. James Flink, in The Automobile Age, his cultural history of cars, stated that “A lifestyle based on mass personal automobility first developed in Southern California, and nowhere in the world has mass motorization been more pervasive in its impact." As early as 1910, the LA region had the most vehicles per capita in the United States. During the 1920s, the population of LA County doubled but the number of registered vehicles increased nearly sixfold, from 141,000 to 777,000. By 1950, the Los Angeles Planning Commission was bragging in its annual report that there were more motor vehicles in LA County than in any other region in the US or than in any other nation on earth.

This rapid adoption of cars and trucks wasn’t without wrinkles and tension. Historian Ashleigh Brilliant discussed how popular and and legal attitudes on mass motorization in the LA region evolved.  What began as a marvel at the new technology turned to shock as residents became exposed to levels of violence from vehicle crashes that few people had ever experienced outside of war zones. Society came to accept the new reality of speeding, death and traffic congestion, partly through black humor.

Faced with a new transportation technology, policy makers scrambled to figure our how to regulate cars on city streets. In LA, early regulations borrowed speed limits from bicycles, parking rules from horses, and traffic procedures from streetcars. By the early 1920s, vehicle proponents and planners saw the need for a more comprehensive set of traffic regulations. In 1925, The city of Los Angeles adopted a new traffic code drafted by Miller McClintock, who helped pioneer both traffic rules and the profession and training of traffic engineering in the United States. As Peter Norton shows in his book Fighting Traffic, this influential code incorporated McClintock’s belief that “the old common law rule that every person, whether on foot or driving, has equal rights in all parts of the roadway must give way before the requirements of modern transportation.” The LA code:

  • set rules for turns at intersections
  • required pedestrians to cross at crosswalks and imposed fines for ‘jaywalkers’
  • banned horse-drawn wagons during rush hour  
  • limited automobiles from dangerously passing streetcars
  • did not prioritize streetcars over vehicles

Other policies, from road dedications to mandatory off-street parking requirements to street design guidelines and speed limits continued to shape a road network and set of rules that were highly favorable to passenger cars and commercial trucks. And while auto dealers and associations helped sponsor and promote pro-car traffic rules, it would be a mistake to think that the public was entirely 'duped’ by a vehicular conspiracy. People liked driving. Jeremiah Axelrod’s history Inventing Autopia recounts how Los Angeles residents protested a downtown on-street parking ban and voted down expansions of rapid transit in the key decades when car-use was exploding in LA.

The take-aways from this capsule history are pretty obvious. 1. New technologies can rapidly reshape how we live and move. 2. Attitudes towards transportation are complex and can evolve to accept what may at first seem controversial. 3. And rules matter in determining what types of mobility predominate.

Keep these in mind when you hear claims about autonomous vehicles. Because the challenges and opportunities are large, I’m happy to see that there are a diversity of visions for an autonomous vehicle future. To over-simplify these viewpoints, I’d classify them into three “schools.” We could call these the robot cars school, the smart streets school, and the tunnels school. These are each worth considering in their own right. I think that they also provide some useful overlapping projections and policy recommendations.

robot cars

The robot cars school posits that improvements to navigation systems (essentially, vehicle software), will allow ‘smart’ vehicles to drive on our existing ‘dumb’ road network. Because the intelligence in the system will be embedded in vehicle software and, presumably, in personal smartphones (or whatever the future version of connected devices are), autonomous vehicles will be able to transform how we get around without the need to reinvent streets or set up centralized systems to guide which cars go where.  

My summary of this approach to autonomous vehicles draws heavily upon a conversation I had with Brad Templeton, a long-time thinker on the topic.  He recently spoke at an event hosted by the Berggruen Institute. Journalist John Markoff, the other speaker, provided a snapshot of how far self-driving car technology has come over the last decade. He has also written about some of the remaining technological and regulatory obstacles to autonomous vehicle development, testing and deployment.

Templeton then cited the internet as an analogy of how he thinks autonomous vehicles might be deployed. Just as the diversity and innovation of the internet rests in individual websites and apps, connected through relatively simple infrastructure, Templeton  stressed that a wide range of autonomous vehicle types could use existing street networks, from small shuttles he calls “the neighborhood elevator” that could take residents to their closest walkable commercial area, to single person robo taxis (half the width of our current cars), to shared autonomous taxis to smart vans. As a user, you wouldn’t really care what you were riding in because they would all be able to get you and drop you off and that waits for transfers between them would be very brief.

While this vision of distributed, ubiquitous self-driving cars rests on a light regulatory touch and doesn’t require much coordination, Templeton does think that governments might have their own version of Waze in the future to address congestion, with capacity limits on certain streets. We also discussed the possibility for bikes to be a part of this ubiquitous network. He suggested that if you wanted to start or stop cycling, small robots could shift bikes around the city, hooking onto a bike and using its own wheels to move it, eliminating the need for a bike share station on every block.

Templeton’s thoughts on autonomous vehicles are laid out in more detail on his website with updates and reactions to new developments at his blog.

While I am mainly interested in how autonomous vehicles might impact mobility and urban form, it is also interesting to consider how self-driving vehicles would reshape the automobile industry and other aspects of car culture. Former General Motors exec Bob Lutz’s essay in Automobile News projects a sense of loss. His title “Kiss the good times goodbye” foreshadows his argument that  “we are approaching the end of the automobile age.” Lutz believes that car as we know them will be replaced by “fully autonomous modules.” Sales to companies in the ride hail and freight businesses will shape the future of vehicles more than traditional marketing of cars to individual consumers. Driving by humans in most circumstances will be banned.  

It is interesting that the robot cars vision of interchangeable modules could presumably both retain small motorized vehicles as the core form of personal mobility and simultaneously undermine the powerful cultural role of car ownership and driving.

Smart streets


current and future streets concept from NACTO

current and future streets concept from NACTO

The National Association of City Transportation Officials, which has emerged in recent years as a sort of complete streets + good urban design counterweight to traditionally-car-and-highway-oriented state transportation agencies and engineering associations, just released a “Blueprint for Autonomous Urbanism”. The preface by Janet Sadik-Khan frames the arrival of autonomous vehicles as  a “historic opportunity to reclaim the street and to correct the mistakes of a century of urban planning.”

The report  “envisions a future where cities and transit agencies leverage new technology as a tool to enhance the public realm and improve the lives of all urban residents.” It includes 6 principles for autonomous urbanism, which I'll try to summarize:
    1.    Safety is the Top Priority = set low (20 mph) speed limits on city streets as the norm and require autonomous vehicles to yield to people.
    2.    Provide Mobility for the Whole City. Assure flexible and affordable mobility options for every community, including walking, biking, fixed transit and ridesharing.
    3.    Rebalance the right of way = Since autonomous vehicles could move more people in fewer cars and less road space, reallocate street space to active and sustainable modes.
    4.    Manage streets in real time = use dynamic pricing and capture data (while protecting user privacy) to manage mobility.
    5.    Move more with fewer vehicles = encourage/ require shared vehicles and nudge public transit to adapt to be a part of the new mix of mobility options.
    6.    Public benefit guides private action = public policy goals should guide autonomous transportation and public/private partnerships.

The report also includes policy goals aimed at implementing these principles. It then suggests (and shows examples of) how different types of streets and intersections could be designed to take advantage of autonomous vehicles and more mobility data. One of the most inspiring sections of the report is a sequence of text and graphics showing how a person in a wheelchair could cross a future street much more easily, because they could A. cross in the middle of block rather than needing to go to an intersection or crosswalk, B. the street wouldn’t have curbs between the sidewalk and roadway, and C. all vehicles would be autonomous and programmed to slow and stop for people walking or rolling in the road.

In contrast to the more distributed, private-sector reliant robotcars model, the smart streets vision embodied in the autonomous urbanism report stresses the need for planning and policy to ensure positive outcomes. Transit should be improved and preserved as the backbone of mobility on high demand corridors. Pricing and regulation of routes, curbside management “or even a passenger trip assignment system” would help prevent competing companies from flooding streets with too many robo taxis.

NACTO’s report can be accessed here.


There is no autonomous cars Tunnels movement per se, but the FAQ of Elon Musk’s Boring Company, a firm that is currently digging a test tunnel under the city of Hawthorne, provides a concise summary of the concept:
“A large network of tunnels many levels deep would fix congestion in any city, no matter how large it grew (just keep adding levels). The key to making this work is increasing tunneling speed and dropping costs by a factor of 10 or more – this is the goal of The Boring Company. Fast to dig, low cost tunnels would also make Hyperloop adoption viable and enable rapid transit across densely populated regions.”

A network of tunnels in dense urban areas would be used to carry individual vehicles (presumably either human driven cars or autonomous cars) on what the boring company call ‘skates’ and considers to be autonomous vehicles in their own right. This would allow faster transportation by avoiding ground level congestion. Longer tunnels would carry intercity hyper loops. The ambition of this tunneling vision is evident in the combination of these two scales: recreating local roads and regional and interstate highways, but underground.     

The viability of these efforts obviously depends on the potential for tunneling to become much cheaper. Many transit advocates and experts I trust bemoan the high costs of transit construction in the United States, but also doubt that tunneling can be radically reinvented in dense areas. Still, it is worth keeping an eye on tunnel options. Shakespeare, Hegel and Marx all praised the revolutionary potential of the “olde Mole,” with Hamlet himself commenting that it was "a worthy pioneer" that “can’st work i’th’earth so fast.”

Robot cars + smart street + tunnels  

These potential autonomous vehicle futures get more interesting when we consider how they might interact. Let’s assume that the key ‘driver’ of future mobility will be robot vehicles and that the intelligence in the system will be distributed in each vehicle rather than contained in a centrally controlled system. However, let’s also assume that we should require that robot cars be programmed to follow low speed limits and defer to people when driving on city streets.  

This combination of the robot car and smart/ complete/ pedestrian-prioritizing streets approach could yield safer local mobility in cities and reduce car ownership and the space needed for vehicle parking. But what about longer trips? As early as Los Angeles’ 1924 Major Traffic Street Plan,  planners were pointing out conflicts between long-distance, high-speed trips across the region and local, stop-and-go trips. It was an early impetus for the construction of grade-separated roads (future freeways and highways) and for off-street parking requirements. Will current freeways remain the fast, long-distance transportation option in the future? Small local robot cars would maneuver safely and slowly among pedestrian, cyclists and other road users until they get to highways, then accelerate.

This is where the tunnels concept seems like a useful alternative. If we want to have a safe, pedestrian-prioritizing city above-ground, maybe most fast travel should occur below. More tunnels could allow more subway transit and/or fast vehicle movement. Some highways could be removed and developed into housing, businesses and open space.

Road fees as part of the mix would help fund conversion of streets to safer designs. (With bike infrastructure and vision zero safety improvements stalling, Los Angeles certainly needs all the help we can get in this regard). Fees could also help subsidize transit (including the ride share shuttle program Metro is exploring) and provide subsidies for lower-income residents to use autonomous vehicles of all types.

I don't know for sure whether autonomous vehicles moving slowly in cities and quickly below is the future we need. But with California and the Federal Government providing a green light  for companies to test autonomous vehicles, it is worth keeping one eye on the history of automobility and the other on options for a post-driving vehicle future.