When Planning for Driverless Systems, Don’t Forget our Values or the Lessons from the Past

May 14, 2018

By Sorin Garber, Sorin Garber & Associates

We’ve been hearing from industry pundits that fully driverless vehicles (i.e., also known as autonomous vehicles, or vehicles that can operate without a human driver under any physical or environmental conditions) will be available for purchase in the year 2021 and will represent 15% of total auto sales by the year 2030[1].[2] Couple that with other predictions that many of us won’t own cars in the 2030s because we’ll be relying on driverless Ubers and transit shuttles that will be in constant motion, and it’s clear we have a lot of unknowns to consider.

There’s disagreement about these predictions, and even more debate about when we’ll realize the full benefits of driverless vehicles (DVs), such as purported lower operating costs, reduced congestion, mobility for those who are unable to drive, and virtually no accidents. But before our imaginations are swept away by the technologies and how they’ll improve our everyday travel, we need to also think through how they’ll affect our lifestyles – including those qualities we don’t want to lose – when planning those investments.  While DVs will bring a great deal of efficiency to our daily lives, we need to ensure they don’t govern how we live at the expense of what we hold dear about our communities.

History of  DVs

Driverless vehicles are not new to the scene. In the mid-1920s, the radio-controlled “American Wonder” traveled on streets in New York City and Milwaukie with other moving traffic without incident. The Futurama exhibit at the 1939 World’s Fair in New York featured a system of automated guided cars and predicted these autonomous vehicles would be a reality by the 1960s.

During the 1950s and 1960s, GM produced the Firebird-series of experimental cars that were described as having an “electronic guide system [that] can rush it over an automatic highway while the driver relaxes”. During the same period of time, the Transport and Road Research Laboratory in the United Kingdom tested a driverless Citroen controlled by magnetic cables that were embedded in the road, and RCA Labs created a test track with detector circuits buried in the pavement which sent impulses to their experimental car. RCA predicted commercialization of their system would be in place by 1975.

In the mid-1980s, Carnegie Mellon University‘s Navigation Laboratory with  Mercedes-Benz and Bundeswehr University Munich‘s Eureka Prometheus Project built the first truly driverless car. Since then, numerous companies and research organizations have developed working prototype DVs. In July 2013, Vislab’s BRAiVE vehicle moved autonomously on a public roadway in Parma, Italy with mixed traffic.

While fully autonomous vehicles are not yet available to the public, many new cars come equipped with  substantial autonomous functions including adaptive cruise control, lane assist, and parking assist.  None of these systems are capable of operating in driverless mode; however, Tesla and Audi recently introduced technologies that combine autonomous steering, braking, and speed limit adjustment under typical and atypical driving conditions, which classifies them as having achieved a Level 2 or 3 on the National Highway Traffic Safety Administration’s (NHTSA) five levels of vehicle automation, where Level 5 represents 100% driverless automation at the highest measure of safety.

In addition to the technological advances, government officials have been promoting the introduction of DVs.  In the early 1960s, the Bureau of Public Roads considered the construction of an experimental electronically controlled highway which attracted the interest of four states – Ohio, Massachusetts, New York, and California – who prepared bids for construction on their highways. During the same decade, United Kingdom officials determined that construction of a fully autonomous system based on communications between electronic cables embedded in pavement and cruise control devices in cars was not only feasible but would pay for itself by the end of the century.

The federal transportation bill in 1991 (ISTEA) instructed the US Department of Transportation (USDOT) to “demonstrate an automated vehicle and highway system by 1997” which was accomplished in a demonstration of close headway platooning using 20 automated cars, trucks and buses on a stretch of I-15 in San Diego. However, the program was cancelled due to budget limitations.

In the past decade, 22 states have passed laws permitting the study and testing of autonomous vehicles.[3] – Oregon is not among them. Overseas, several major European cities are planning to operate driverless transit services, and Germany, the Netherlands, and Spain have allowed testing DVs in traffic.

Back to Planning: Will We Need More or Less Roadway Capacity?

If we do build (or allow others to build) the systems to allow for fully operational DVs, in theory, we won’t need as much roadway space to accommodate traffic. The research on highway capacity maintains that if we could all drive at a constant rate of 60 mph, our highways could move about 2,000 to 2,400 vehicles per hour per lane; which is largely based on the spacing between vehicles and the time motorists need to observe and respond to stimuli such as a stopped vehicle.

Under a fully connected driverless system – where we can reduce the distances between cars because the systems will be able to react to stimuli faster than humans – simulations predict that we’ll be able to accommodate 8.000 vehicles/hour/lane, or just 25%-30% the space we provide today. That means we might not have to add many or (in some cases), any more lanes to highways.  We might even be able to eliminate some lanes to build parks or natural areas or housing, etc. But then what happens when those 8,000 vehicles/hour/lane leave the highway at the exit ramps and travel onto downtown streets or local residential roads?

Travel on local streets could also be automated to meter traffic flows, slow speeds, and control stopping for vehicle drop-offs and pickups.  In addition, the DVs may be able to drop us off and then park themselves elsewhere until we call them up to come pick us up, so parking near your destination won’t be necessary. Our downtowns also won’t require as much roadway space for our mobility.

However, there are at least three arguments for why we may need more roadway capacity in the future of connected and driverless vehicles. First, we’ve found that whenever the cost of driving drops and congestion lessens (both of which have been predicted as benefits of DVs), we usually end up driving more and thus require more roadway capacity. Add to that, the anticipated driverless and continuously circulating ride-sharing vehicles (such as Uber, Lyft, and other transportation network companies), which not only add more traffic to our streets, but also divert passengers from our transit systems which presently help to reduce the number of vehicles on our streets. Then there’s my greatest worry, that we may find that long commutes are acceptable because we’ll be able to accomplish multiple things while the car is driving us, and thus require added roadway capacity to accommodate a new urban sprawl trend.

Timing is Everything

Public investments in our transportation infrastructure and systems are planned in 20-year cycles which we revisit and refine every four to six years.  These cycles have worked in the past because we’ve generally had a good grasp about how to forecast future travel, and thus, plan, design and construct systems that meet our needs. Many of our large transportation projects – from new bridges to new light rail transit systems – take as long as 10-12 years to complete to account for the formation of financial partnerships, completion of design, acquisition of needed land and regulatory permits, and construction. Even those projects that are “accelerated” through design build methods only save 2-5 years total, because much of the time required to implement projects involves the financing elements and their associated unpredictable public processes, legal and policy reviews, and political agreements.

If it takes 12 years to build new projects and DVs will penetrate the market in significant numbers in 12 to 17 years, we run the risk of designing facilities and systems that don’t synch up with our future use of DVs.

Also, if DVs can improve access to all parts of our region and at a lower cost, how will that affect the demand for transit services?  DVs, particularly those operated by transportation network companies, could divert some travelers from transit services where they can provide a lower cost, improved comfort, and/or faster travel time than conventional transit. Perhaps transit will replicate these benefits and also use driverless technology, such as the driverless transit shuttles currently operating in Paris, Singapore and Arlington, TX.

Should we do nothing and simply reap all the benefits of DVs? Or how do we make the best investments?

No one questions that we need to start thinking about the infrastructure and other system and service investments needed to accompany DVs, particularly when or if they conflict with our public infrastructure investments, as well as if they appear to have the potential to overburden them. Many jurisdictions are on the right track in thinking about regulating DVs, particularly curbing the continuous operations of transportation network carriers (TNCs), in large part, because they are most aggressively developing DVs to operate in the very near future.

Very few jurisdictions, however, have had serious and rational discussions about what systems are needed to accommodate future travel because some of the DV operations are truly revolutionary and we’re unclear what to do about them. But doing nothing will result in our later having to react to what DVs have brought us – such as what we did when we relinquished our transit systems in favor of the automobile.

Forecasting our future needs is complicated even when we have a basic understanding of the changes we can reasonably expect, such as population growth, demographics, land types and availability, and cost of living.  Imagine, how much more complex it is to forecast our needs based on unknown factors such as continuously operating DVs, systems that have intermittent connectivity, declining auto ownership, decreasing revenues for roads and transit, and driverless trucks?

While there will be intermediate stages before DVs dominate our roadways and we’ll need to keep our conventional systems in place, we do need to be thinking long-term toward when they are the dominant vehicle type. This will create new approaches to our planning. For example, our forecasts need to recognize that:

• There will be many more travel options for all types of trips and these different options will be competing for our patronage.
• Travel by transit, ridesharing or TNC services will be extremely convenient and low-cost, and each system will be interconnected.
• DVs may be able to replace some of our paratransit services.
• Congestion levels will decrease in many facilities due to instant communications about upstream congestion and quick mobilization of detour routes and/or options.
• DVs will be incorporated into services provided by retailers, restaurants, tourist destinations, educational and training institutions, and others.
• Traffic safety will improve significantly because of the automated braking, steering, and navigation systems in DVs.
• Much of the essential support systems – communications, GPS, etc. – will be the responsibility of the private sector who will be reimbursed for those

Let’s not get Swept Away by the Whiz Bang Technology

Driverless technology has the potential to transform cities in many positive ways. In time, DVs will accelerate our ability to better connect places we live, work at, visit, shop, recreate, worship, etc. Eventually, they will allow those who are unable to drive to have another option for travel. They will virtually eliminate traffic accidents, and they will significantly reduce carbon emissions and fuel consumption. They will also have an unprecedented positive impact on our economy. Frankly, it’s hard to imagine all of the different ways that DVs will change our daily habits and influence our decision-making.

But we can’t let the technologies govern the conversation about our future needs because we’re unsure exactly what the technologies will bring us. We need to be proactive about what we want from the technologies (e.g., improved safety, mobility for those who can’t drive, lower costs, reduced emissions and fuel consumption) and not have our decision-making dominated by the immediate, and potentially short-term, conveniences of DVs (such as improved travel time, lower costs, their ability to park themselves, etc.).

If we do nothing, DVs could end up increasing congestion, reducing transit ridership and revenues for transportation improvements, encouraging sprawl, and exacerbating inequalities between those with the means for the DVs and those without. In other words, we need to establish public policies that shape DV technology in ways that preserve and enhance the quality of our communities and lives of our residents and workers.

There may be a few initial investments for the next 20 years we can all agree on – e.g., establish dedicated lanes on freeways and other principal roadways for driverless HOVs (e.g., carpools, buses, vanpools, taxis) and driverless trucks, so they can be platooned and use less roadway space.  And, maybe 20 years from now, we’ll have figured out how to then implement connected systems for DVs on local streets and in commercial centers. But as we think about these possible futures, we need to put away our Jetsons’ imaginations, and remember what we want our future world to be.

---

Sorin Garber (Sorin Garber & Associates) is a transportation planner in Portland who has prepared growth forecasts and analyzed the effects of new transportation technologies and facilities on roadway congestion, transit services, and goods movement throughout the country for over three decades. He can be reached at sorin.garber@sgapdx.com

---

[1] Vancouver’s Mobility Future: “Automating” Policy into Sustainable Results, presentation to Vancouver City Council by Dale Bracewell, January 30, 2018, https://www.scribd.com/document/370838534/Vancouver-and-AVs-With-Transit

[2] Autonomous Vehicle Implementation Predictions, Implications for Transport Planning, Todd Littman, Victoria Transportation Institute, December 22, 2017 http://www.vtpi.org/avip.pdf

[3] Autonomous Vehicles/Self Driving Vehicles Enacted Legislation, National Conference of State Legislatures, 3/26/2018 http://www.ncsl.org/research/transportation/autonomous-vehicles-self-driving-vehicles-enacted-legislation.aspx