Impact of Self-Driving Cars Michael Morscher


Paving the Way for Infrastructure Evolution

Everyone knows America’s infrastructure desperately needs improvement. Most people know that America’s infrastructure as it stands is insufficient to support autonomous vehicles. Those who don’t know would probably be able to guess correctly.

But, in this plight lies an opportunity: the opportunity to be proactive, making the necessary investments in infrastructure improvement and doing so with self-driving cars explicitly in mind. We have seen light- and heavy-rail transit systems across the country have their service deteriorate as their infrastructure deteriorates. Taking the MTA and MBTA for example, what becomes clear is that the cost to repair infrastructure skyrockets the longer it is neglected. Public perception of this matter is understanding, since rails and signaling systems are something tangible. Yet even though people are fed up with poor road conditions, there is a tendency for public agencies to do cheap and lazy repairs like pouring tar in potholes and hoping it works out, and many accept this as “good enough”. But this mindset is not good enough. Resolution of underlying problems and preemptively fixing anticipated problems is the only way infrastructure will be able to be maintained for AVs, and since they will end up making roads more like integrated rail networks with signal systems, a lesson can be learned as to when this investment ought to be made.

There are persistent (albeit ineffectual) rumblings from both sides of the political aisle about infrastructure improvement, and this fall we witnessed bipartisan agreement on regulating self-driving cars. Rather than looking to historical inaction, there is reason to remain optimistic that legislators will recognize this need and act responsibly.

So, let’s recap the considerations that will need to be addressed.

In industry, there are three lines of thought regarding the infrastructure changes needed to accommodate self-driving cars. In truth, it is almost surely a combination of the three that will be most socially beneficial:

  • New infrastructure. The most expensive of the three categories, but unavoidable to make the necessary technological adaptations.
  • Rethinking existing infrastructure. Not only repairing, but repurposing.
  • Removing infrastructure. It is no one’s first instinct, but we need to be proactive about current infrastructure that is sliding into obsolescence in order to be ready for the AV revolution.

From my research I believe rethinking existing infrastructure holds the most promise, both in keeping costs as low as possible and facing the fewest barriers to progress. Moreover, it is the most reasonable place to start while we still have an overwhelming majority of traditional cars in operation.

Within this, the three overarching motivations seem to be:

  • Redesigning cities around the pedestrian
  • Redesigning highways around the self-driving car
  • Redesigning roads around both

In each, there are conflicting priorities and debate regarding changes to be made. The least debate probably lies in the highway–there really is no particular reason to make it a pedestrian paradise. If we can accept this as a fact, then it shouldn’t be a stretch to recognize that there’s no particular reason to optimize cities for cars rather than pedestrians. Urban roads will inevitably require compromise, but the compromise is probably easier to attain outside of cities as a starting point.

At the end of the day, wouldn’t it be nice to live and work in a city that looks like this?

The changing landscape of city streets as a result of self-driving cars.

There just might be infinitely many factors to consider when redesigning what the world we live in looks like. It can be confusing and intimidating. And at the end of the day, the people who will be tasked with making these decisions are not the people who are experts in the field. I’m certainly no expert myself, but to reduce the management of the potential impact of self-driving cars to as simple an initiative as possible I say: keep it realistic and responsible.






Filed under: Uncategorized No Comments

A Disappearing Act

Not all of the impact on infrastructure resulting from the deployment of self-driving cars will be innovative high-tech replacements of the infrastructure we already rely on. A great deal of the change will likely be today’s infrastructure disappearing.

One of the articles in the recent New York Times Magazine issue dedicated to self-driving cars is titled “Cities Without Signs“. It makes the point that without humans needing to navigate in their cars, the concept of needing street signs at every waypoint is obsolete. AVs can direct themselves through GPS data, control their speed without speed limit signs, and safely operate without abiding by stop signs, “do not enter” signs, and yield signs. Why would we build and maintain costly signage that goes unused?

“…No huge billboards across highways naming the exits, no complex merge instructions. Certainly, human pedestrians and cyclists will still need guideposts, but as Stone suspects, far fewer, and smaller, ones… It could make a city less cluttered and more attractive. But it might also leave us feeling unmoored.” 

Indeed, for those of us living our whole lives with omnipresent infrastructure features, it will feel strange seeing them disappear. Future generations will only read about street signs and these other artifacts in history books:

AV adoption on a shared-use model would let every home garage be transformed into a new space.

  •  Home garages. Who would dedicate 20% of their living space to a car they don’t own and parks itself at remote locations for the few hours of a day it even needs to be parked?
  • Parking lots and public garages. Same considerations, but on an even larger scale. Land that is made available can be used for massive urban development projects rather than just home renovation.

Parking garages can be gradually reclaimed and eventually eliminated with the utility of self-driving cars.

  • Mechanic shops. Regardless of how much auto repair itself becomes automated, it certainly won’t require prime real estate if cars take themselves in for maintenance.
  • Traffic lights. In the short term with partial AV adoption they will become even more essential, but when all cars become autonomous they can be eliminated.

AVs will be able to use virtual renderings of intersections instead of traffic lights.

  • Gas stations. AVs and EVs go hand in hand, making gas stations a thing of the past…

“There might be a time in the far future where you explain to your grandchildren about how you had to refuel your car — that wasn’t self-driving — at a special place in town. Maybe they’ll giggle thinking about the ridiculousness of putting a liquid in a car. But when you tell them about the selection of beef jerky at this “gas station” their eyes will grow wide, which will be followed by complete disgust as you describe the filthiest bathroom you’ve ever encountered.”  (Thank you Engadget for playing perfectly into my narrative)

We are only at the early stages of autonomous technology. Many other facets of infrastructure will unexpectedly become obsolete. And they will disappear.

But from this exodus will arise a new opportunity in infrastructure.

The possibilities of how to use all this reclaimed land area and air space are endless. What would you do with it?



(“Self-Driving Cars Will Kill ________” is an incredibly popular headline)

Filed under: Uncategorized 2 Comments

Starting from Scratch

With all the problems of adapting existing infrastructure for compatibility with autonomous vehicles, wouldn’t it be nice to be able to start from scratch and design cities from the ground up with AVs in mind? However exciting this would be, I can’t envision tearing down and completely redeveloping the cities we live in today. It’s conceivable enough that bite-sized models of cities can be constructed as testing grounds to simulate real existing cities, but what about new urban communities people can actually live in? I was surprised to learn of the scale of new cities that are still under development today.

Castle, CA. Waymo’s artificial city for testing self-driving cars. Check out the cozy homes those driveways lead to…

One such city is Babcock Ranch, Florida. Babcock Ranch was planned to take full advantage of future technologies such as complete solar power dependence and (of course) self-driving cars. What’s astonishing is the ease of development they are facing in this endeavor—since they are building from scratch they face no legacy issues from existing infrastructure, and since they are privately owned and controlled they face no bureaucracy issues.

The implications of these factors are also very interesting to consider and eliminate many of the traditional concerns of urban AV deployment. Babcock Ranch reserves the ability to restrict private car traffic which would prevent predicted conflicts, and limit coordination to the much more manageable use case of cyclists and pedestrians. Since people are buying in to this new community, they know what they’re signing up for and there won’t be a pubilc resistance to restricting car use. In fact, this quality might actually be a draw for the hipster types attracted to new age exclusive urban development!

A self-driving shuttle in the newly constructed city of Babcock Ranch, FL. Now this is a city!

Other implications include the ability to easily change the city around them as the technology catches on, redeveloping parking lots into parks with no opposition. It is also intended to be a proving ground for new concepts like automated package delivery within the city. All of this and more made possible by starting from scratch and designing for change.

“What we’re hoping people realize when they move to Babcock Ranch is very quickly they’re only going to need one car, and then our ultimate goal is for them to realize they don’t need any cars. Think about what that means from a perspective of the home buyer–they now have a two-car garage that could be used for something else.” – Syd Kitson, CEO of the company developing Babcock Ranch

Babcock Ranch was designed with technology and adaptation at its core.

Side Note: Another hidden benefit of forward-thinking urban planning: in contrast to the nearby planned city of Cape Coral which was built in the late 1950s, Babcock Ranch survived recent Hurricane Irma essentially unscathed.

Even though our minds are often closed to consider only the cities we already inhabit today, there are hundreds of new cities currently under development (including many in Florida and California) with the opportunity to deploy revolutionary designs far sooner than existing cities ever could. According to analysis by the New Cities Foundation one of the most important factors in urban design is the ability to adapt to new technologies and forms of transportation. It’s becoming increasingly clear that self-driving cars will be the largest and most important societal change to come in the near future, and I would strongly encourage developers of emerging communities to seize the opportunity and plan with AVs in mind.

As many different ideas about AV infrastructure as there are, the National Association of City Transportation Officials released the comprehensive Blueprint for Autonomous Urbanism this week, and it might just become a gold standard for urban planners to reference in this scenario.

One of many NACTO renderings for sustainable urban AV infrastructure, to be used as a reference by urban planners.

For the sake of actually seeing new infrastructure in action, and for having a proof of concept before we remodel existing cities, we should look to the cities being built from scratch for inspiration and guidance. At the beginning they will likely be forced to be self-contained, but over time they have the potential to integrate with surrounding cities and infrastructure to drive the self-driving revolution.



Filed under: Uncategorized 2 Comments

Autonomous Vehicles on City Streets

As promised, I will be evaluating the efficacy of dedicated lanes or streets for autonomous vehicles within cities. It’s obvious that city roadways have many special considerations such as intersections and mixed traffic usage, but even urban highways differ from rural highways, with a greater frequency of exits and entrances and more complicated traffic interchanges. These factors lead to two contrasting conclusions: that grade-separation of AVs is significantly more important in the urban setting, and that it would be significantly more difficult to implement.

GM’s “Futurama” exhibit at the 1939 New York World’s Fair showing grade-separated city streets.

One recent proposal known as Loop NYC would reserve one lane in each direction (with assumed unlimited throughput) on the highways surrounding Manhattan, as well as some of the major crosstown streets to move AVs within the city. This would seemingly create a perfect balance between AV optimization and preserving regular traffic with minimal obstruction, additionally creating park space for pedestrians. Yet, pedestrians would unavoidably and inconveniently be forced onto overpasses to cross this network, a prospect inconceivable for influential urban planners.

Loop NYC’s rendering of proposed AV lane network in Manhattan

Similarly in Atlanta, plans were made early this year to designate a “smart corridor” as a proving ground for AVs, but there were persisting concerns about pedestrian safety particularly near the Georgia Tech campus where students are known to jaywalk. For the time being, the corridor remains a road with intersections and stoplights rather than a converted highway, reducing pedestrian conflicts but tanking AV efficiency. About a month ago autonomous buses were launched on this roadway “without catastrophe.”

In states including Nevada and Ohio, emphasis is being placed on deploying “smart roads” with integrated sensors to enable and improve self-driving performance. Having dedicated rights-of-way for this technology helps make it practical and optimizes its use. Especially in dense urban areas, V2I and V2V (vehicle-to-infrastructure and -vehicle) technology is essential for safety and coordination, and these technologies perform best in constrained and exclusive settings.

The Loop NYC project was one of seventeen entries in the Driverless Future Challenge, but the concerns for pedestrian experience prevented it from being selected as the winner. To put their shortcoming in perspective, the winner (Public Square) was concerned exclusively with the urban pedestrian experience. Their mission is to “reclaim the street”, and as AVs reduce traffic, they plan to gradually convert parking spaces into parklets until entire streets become linear parks.

This challenge in infrastructure planning is particularly challenging because there is no innovative solution that is fair to all forms of transportation. Individual forms can be completely optimized, but at the great expense of other forms, and there exists no possible optimization of all current and future transit options. The takeaway:

“As long as they share the road with pedestrians, bikes, and human-driven vehicles, self-driving cars will not be able to reach their full utility. The question is, what would cities have to sacrifice to unlock that utility?” Benjamin Schneider, CityLab

Architect Harvey Wiley Corbett’s vision for a grade-separated city in 1913.


GM created a sequel to the “Futurama” exhibit for the 1964 New York World’s Fair. Included below is an incredible video of the entire exhibit, but their rendering of a future city at the end is definitely the highlight. They predict, “Plazas of urban living rise over freeways. Vehicles, electronically paced, travel routes remarkably safe, swift, and efficient. Towering terminals serve sections of the city, make public transportation more convenient, provide ample space for private cars, and from a lower level, covered moving walks radiate to shopping areas that are now truly marketplaces of the world.”

Self-driving cars and grade-separated cities in 1964? I’ll put my money on GM to be the one to make it happen in 2017.



Filed under: Uncategorized 1 Comment

HOV Lanes. EV Lanes. AV Lanes?

Human drivers sharing the roads with autonomous vehicles is assuredly problematic. And, since the government can’t even get potholes repaired, it’s looking like we won’t be able to construct an entire parallel network of roads exclusively for self-driving cars in the foreseeable future. There’s going to have to be a compromise in the coming years which takes advantage of existing infrastructure and allowing new technology to exist with the old. About 100 years ago, drivers of the Ford Model T had to battle with horse-drawn carriages until eventually the transportation advancement became ubiquitous. This problem is not unlike that faced today, which leads to the question: If we already have dedicated lanes for HOVs (high-occupancy vehicles) and EVs (electric vehicles), why not dedicate AV lanes?

Are dedicated highway lanes for autonomous vehicles a natural extension of HOV and EV lanes? Many industry leaders say ‘yes’.

One recent proposal asks for just this. Autonomous vehicles would be allowed in existing HOV and EV lanes by default, with a gradual phasing out of existing traffic from these lanes—and soon from the entire highway. This plan would implement an immediate solution for the small amount of AV traffic out there right now and dynamically adapt to decreasing levels of human traffic.

By 2025, that HOV lane would be closed to human drivers, along with the one next to it (the highway is eight lanes wide)… those two lanes could be used as three, as autonomous vehicles can safely drive more closely together. By 2030, the majority of the highway would be closed to human drivers, with the takeover complete by 2040.” – Wired

Another proposal to designate AV lanes known as Hyperlane would integrate sensor systems with a central computer to monitor traffic volume and optimize travel speeds. There is broad industry recognition of the idea that “smart roads” which communicate with cars, and allow cars to communicate with each other, are essential to optimize speed and safety. But there is a multi-trillion dollar price tag. Deploying this only to select lanes would be a great short-term solution to provide necessary technology while controlling its cost.

“Hyperlane” is a proposed network of highway sensors that would allow self-driving cars to synchronize and reach speeds in excess of 100 miles per hour.


Existing highway lanes are rarely in ideal condition, which poses a major problem for current low-level AVs that are dependent on lane sensing. Self-driving cars are programmed to stop or return control to a human driver if input is insufficient, which is what happened during a Volvo demonstration at the Los Angeles Auto Show.  Their North American CEO exclaimed,

It can’t find the lane markings!… You need to paint the bloody roads here!”. – Lex Kerssemakers, and probably echoed by all other drivers in America

In fact, about two thirds of American roads are in poor condition, and auto makers have explored self-sufficient technologies but in the end infrastructure must be made compatible. Again, having dedicated lanes in this scenario would allow for focused investments in lanes where cars can’t operate without clarity.

“Smart roads” would communicate with autonomous vehicles to increase efficiency of traffic flow and ensure safety.

There is always the opportunity for traditional drivers to feel neglected if AVs are prioritized or receive dedicated funding, but AV lanes would strike a good balance in this tradeoff. If existing restricted lanes are used, traditional drivers won’t have fewer lanes than they do now (for the time being) and people converting to self-driving cars would reduce overall traffic. Targeted investments in infrastructure or technology might however burden taxpayers who wouldn’t benefit from the advancements, which is definitely a concern to consider, but differential tolling on these lanes could offset or eliminate this inequality. If the downsides of AV lanes can be overcome, the potential benefits are unbounded and would provide a clear road map for full AV adoption.

Can this same principle be applied in the urban setting with grade-separated roadways? Or are the special considerations of cities such as shared traffic just too difficult to overcome? These questions and more will be the focus of my next blog post.



Bonus Reading:

Amazon thinks it can handle the rarely considered scenario of reversible highway lanes:

Filed under: Uncategorized 2 Comments

Don’t Get Your Signals Crossed

Whenever it is that self-driving cars begin to roll out to the mass market, there will be a period of time when traditional and autonomous vehicles share the road together. In certain applications such as highway driving, there probably won’t be too much of a clash. We can look to the presence of cars such as the Tesla Model 3 currently on highways today, which for all intents and purposes in this narrow use case, behave just the same as higher-level automation would. However, one clear case in which there could literally be a clash between self- and human-driven cars is intersections.

We can imagine what intersections would be like with 100% of vehicles automated, with cars zipping past each other with centimeters to spare (check out these cool simulation videos).  But with unpredictable traditional cars in the mix, this cannot become a reality. One short-term solution to this problem that can be implemented with minimal infrastructure change is the use of “priority traffic signals”—traditional stoplights that communicate with vehicles deemed to be of high priority and are accordingly granted green lights.

Active priority traffic signals are already used in some cities to ensure green lights for bus rapid transit.

Priority signals have already been implemented around the world for use along bus rapid transit routes without a dedicated right-of-way, having to share the road with vehicles. Take for example the gradual deployments of priority signals to the Silver Line in Boston and Select Bus Service in New York City. Of course this prioritization comes at the expense of people who have to wait a few extra seconds in their cars, but to the benefit of the many more people that can fit on a bus.

Surprisingly, there hasn’t been much of anything published regarding the extension of priority signals to self-driving cars. Any analysis out there escalates immediately to the futuristic hyper-speed intersections that make for those cool simulation videos. Yet, priority signals would allow higher vehicle throughput by prioritizing cars that are able to safely go faster and coordinate to use more intersection surface area at a given time, reducing traffic jams for everyone. And, it would also help obviate the problem of autonomous vehicles battling with human drivers.

As outlined by the Federal Transit Administration, there are two types of priority traffic signals. “Active priority” occurs when a transit vehicle communicates with a light to ensure it is green when arriving at an intersection, as in the image above. “Passive priority” however is easier to implement, and simply refers to setting light timers to be green more frequently for continuous movement of slower traffic, such as buses. Take this visual for example, showing a smoother flow of prioritized traffic.

Passive priority signals manipulate the timing of traffic lights to favor continuous movement of “prioritized traffic” with a different average speed than other traffic.

Theoretically, passive priority could work for autonomous cars if timers are programmed to favor faster traffic, but active priority signals taking advantage of the autonomous technology is the optimal solution.

There is one final concept in existing priority signal setups that would be highly extensible to self-driving cars: queue jumping lanes. This infrastructure modification at intersections creates a dedicated lane with dedicated traffic signals to prioritize eligible vehicles at a red light before other vehicles can start moving again. Creating queue jumpers for self-driving cars would allow them to get ahead of traffic and reduce battling with traditional cars.

Queue jumping lanes used in coordination with priority signals help promote traffic throughput.

These simple retrofits of existing infrastructure would enhance the flow of mixed traffic in a theoretical sense, and perhaps make it possible at all in a practical sense. Best of all, the costs to implement this concept would be an affordable alternative to the inconceivably immense costs to build an entire parallel infrastructure only for self-driving cars. A DOT estimate from 2002  claims the cost is as low as $8000 per intersection, quite a reasonable distributed cost for the trillions (yes, trillions) of miles driven in the US every year.

So, what do you think? Should the industry visionaries and hype men take notice of the “here and now” and explore priority traffic signals as a means to permit mixed traffic? Or should they keep dreaming, hoping that this problem will resolve itself in time, and leaving us to deal with the mess when it doesn’t? I think I’ve made my point to them clear: Don’t get your signals crossed.

News: A few weeks old at this point, but tangentially related. Who would have ever thought hacking a self-driving car could be done with stickers on traffic signs?

Filed under: Uncategorized 1 Comment

The Future

For quite a few years, the concept of self-driving cars has been a matter of science fiction and written off by the general public as achievable only in the distant future. However, the “future” is now here. Cars with low-level automation such as lane keeping and Tesla’s Autopilot are available on the consumer market, and various companies are testing advanced autonomous vehicles on public roads.

Just this week, a paper was published by leading researchers comprising of definitive guidelines based in psychological principle for convincing society to embrace the proliferation of self-driving cars. This was in response to the finding that this year, 3 of 4 Americans are still uncomfortable with the notion of autonomous vehicles (AAA). The findings of the paper are summarized in this article, and it’s worth a read as an introduction to many of the frequent questions that this advancement in technology raises.

While the subject of how self-driving cars themselves will be designed is open to anyone’s imagination (see this creative new design unveiled by Ford today), this blog will analyze a less considered topic: the impact of self-driving cars on the world around us. Surely, today’s infrastructure will need to be modified–and in some cases completely rebuilt–to accommodate autonomous vehicles and optimize their performance and safety. What exactly those changes will be is much less clear. I plan to look to the changes resulting from past advancements in transportation as well as to the forefront of today’s research to synchronize the infrastructure changes we can expect as fully-autonomous cars are brought to market.


News: A lot of headlines breaking today about the rush to be the first and best self-driving car company to market.
Volkswagen joins the list of essentially every existing car manufacturer promising an autonomous fleet within five years (see here), while Samsung (here) and Google (here) announce new partnerships in their respective superiority quests.

Bonus Content: This classic scene from Silicon Valley. Enjoy!



Filed under: Uncategorized No Comments