Self-driving vehicles and the future of parking

    The anticipated marriage of ride-sharing services such as Uber and Lyft with self-driving vehicles should be a game changer for transportation.

    Many people, particularly engineers, policymakers, and urban planners, are anxiously awaiting the introduction of self-driving vehicles (SDVs). They promise to make our streets and highways safer by eliminating driver error.

    However, as promising as SDVs are, they also will bring challenges. Perhaps no segment of the infrastructure will be impacted as much as parking. Why parking? Because personal vehicle transportation is so inefficient. Multiple studies have found that cars sit parked 94 percent of the time. In addition, parking typically requires as much, if not more, floor area as the commercial destination it services. As commuters and shoppers increasingly rely on SDVs that can drop them off and move on, the need for parking facilities will decrease dramatically.

    Even though SDVs are still a few years away, demographic changes are already reducing parking demand. There has been a shift in the attitudes of Millennials (and presumably future generations) about driving and owning cars. Millennials prefer urban living to a greater degree than earlier generations, which understandably decreases their need to own personal vehicles.

    While many will no doubt move to the suburbs when they have school age children, others will choose to stay urban, and still others will likely choose denser walkable neighborhoods near suburban downtowns. Through 2016, vehicle miles traveled per capita are at the level of 1998; and there’s no certainty that we will ever get back to the peak 2005 levels.

    Obviously, it will take time for older cars without the self-driving technology to be retired from service, and it will also take a while for SDV technology to be trusted and widely used. However, the anticipated marriage of ride-sharing services (such as Uber and Lyft) with SDVs should be the game changer for transportation — not to mention parking demand. In fact, it is now expected that public-private partnerships (P3s) will be developed city-by-city, to provide:

    • the infrastructure needed for SDVs (everything from the cellular data network to mapping to maintaining pavement markings), and
    • a local subscription ride-sharing service that will not only supplement transit services in off hours but solve the first mile/last mile problem for transit as well as provide mobility for the aging and disabled.

    In essence, it will be similar to the development of the electric grid in the late 1800s, starting in the downtown of urban areas and radiating outward.

    Parking designers must create sufficient drop-off areas for people who arrive via self-driving vehicles, whether they send the vehicle to park onsite or it goes on its merry way.

    How much could SDVs affect parking demand? Two credible studies used detailed travel data from the National Household Travel Survey (NHTS) collected by the U.S. Department of Transportation to estimate the maximum potential reductions in vehicle ownership if all households that could use SDVs did use them. Using this data, the University of Michigan estimated that nationwide, American automobile ownership could decline from 2.1 vehicles per household (today’s average) to 1.2 vehicles per household. That is a potential decline of 43 percent, simply because one SDV could handle all household trips (that is, without any subscription service). The reduction will certainly vary by area, with urban downtown vehicle ownership dropping by much more than that in rural areas.

    Using the Ann Arbor, Mich., data from NHTS, Columbia University estimated that 18,000 subscription SDVs could provide the trips now made by 120,000 residents of Ann Arbor, which represents a potential reduction of nearly 50 percent in the 200,000 vehicles now owned by Ann Arbor residents.

    While the study’s authors didn’t calculate a reduction in parking demand, it is possible to use their data to extrapolate how much parking demand would reduce. Adding a reported 50,000 average daily commuters from outside the city limits, and assuming that all non-residents won’t be able to use subscription services to commute, we estimate that Ann Arbor would experience a reduction of about 40 percent in parking demand. This is solely due to the subscription model, and not owners choosing to have their own SDV. The two studies thus overlap, but an overall maximum reduction of 40 percent in parking demand seems a reasonable projection of the maximum impact, from what we know today.

    Considering that the typical parking structure is designed to remain in service for 50 years or longer, developers and their parking designers need to address self-driving vehicle-related design issues today, even though SDVs won’t be common for another decade or longer.

    It’s easy to be skeptical about the ability of any single technology to transform the transportation and parking landscape. However, SDVs are different because they have the strong support of the government, regulators, and academics because they offer significant public safety benefits. A study of crashes by the National Highway Traffic Safety Administration found that the most critical factor in 94 percent of crashes was driver error, with the remaining 6 percent caused by vehicle component failure, environment, and other unknown causes. SDVs are essentially a life-safety issue. At some point, those who don’t have the vehicle drive itself may be considered as risky as drunk drivers.

    Parking will also become more efficient. By 2025, it is expected that most new cars sold will be able to drive and park without human intervention. Many manufacturers already offer some form of parking assistance, and in 2015 Mercedes Benz became the first manufacturer to offer fully autonomous parking — the car can find a stall and park itself after dropping off the driver and passengers at the front door. That will allow four cars to be parked in three stalls, primarily because space for opening car doors will not be required at the parking stall. In turn, that represents up to a 1/3 increase in parking capacity, at the same time that other users no longer park at the facility at all.

    While 2025 may seem a long time away, designers of parking facilities need to start recognizing these issues today. The typical parking structure is designed to remain in service for 50 years or longer. The financial markets are already beginning to question the long-term viability of parking facility bonds or loans with terms of 20 or 30 years.

    Some in the planning community are advocating all new facilities be “adaptable” structures that can be largely if not entirely converted to future uses. However, I don’t think that is necessary, nor advisable for many new parking structures. When existing parking lots or structures already serve a destination, such as in a downtown or a campus, the first step will be to develop new uses on surface lots. New buildings will be built without as much, if any, parking. And if a facility is no longer needed, it will be far more appropriate to tear down an old, deteriorating garage than to convert a new, state-of-the-art one. The reduction in parking demand in downtowns will happen over a long period of time and be absorbed into the local parking market.

    Walker Parking Consultants has designed a fair number of parking structures for future expansion during the last 50 years, and I doubt that 1 in 100 was ever expanded. The cost to design a new garage for full conversion to a specific future use is relatively high — as much as a 20 percent premium — and there is no guarantee that the provisions — and investment — will be appropriate for the new use at the time of conversion.

    Building codes change. For example, after ADA was passed, the required clear width between handrails in stairs was increased to allow firefighters to carry people in wheelchairs down stairs. Seemingly minor today, a future code change like that could significantly escalate the cost of conversion.

    However, there are many cost-effective provisions that can be incorporated in designs, including designing the grade and roof levels for conversion, leaving room for the vertical pedestrian circulation and stairs outside the structure footprint. The site can also be designed to allow parking structures to be wrapped with future residential developments. And other provisions such as providing extra capacity in the foundations, additional floor-to-floor height, and using flat floors with express ramps are reasonably cost effective (maybe a few percent greater construction cost), leaving the future structural, MEP, and architectural needs to be designed and retrofitted at the time of conversion for the then-proposed use.

    Another idea is to be more aggressive with parking management now. For example, for residential parking, use car stackers for the second car owned by a tenant and build only one slightly taller level of parking rather than two floors. The tenant pays a lower rate for the second stall to compensate for having to move one car to get to the other. In the future, if the household doesn’t have a second car, the upper pallet can be converted to a storage unit for that tenant and thus still generate revenue.

    An important issue for designers today will be to create sufficient drop-off areas for people who arrive via SDVs, whether they send the vehicle to park onsite or it goes on its merry way. We suspect that a fair number of on-street spaces will be converted to passenger loading zones for car-sharing services such as Uber and Lyft, but in many cases, having the drop-off inside the parking structure may be appropriate and best planned today.

    The reduction in parking demand resulting from SDVs will also change the way we view parking. For nearly a century, American planners and political leaders have believed that, when it comes to parking, more is better. This attitude has already begun to change, but it is to be hoped that more planners and municipal leaders will realize that it makes more sense to build only as much parking as you need to help commerce thrive.

    During the next two decades, SDVs will transform the parking landscape and change the way we as engineers design parking facilities. For municipal planners, city leaders, and private developers, now is the time to begin planning for the new parking reality.

    Smart Growth America releases transit-oriented development parking study

    Hundreds or thousands of people travel to and through transit stations each day, and decisions about what to do with the adjacent land have implications for local economies, transit ridership, residents’ access to opportunity, and overall quality of life for everyone in a community. Many communities choose to dedicate at least some of that land for parking. The question is, how much? Standard engineering guidelines are designed for mostly isolated suburban land uses — not walkable, urban places served by transit. But few alternative guidelines for engineers exist.

    Empty Spaces: Real parking needs at five TODs, set out to determine how much less parking is required at transit-oriented developments (TODs) and how many fewer vehicle trips are generated than standard industry estimates.

    Professor Reid Ewing and his research team at the University of Utah College of Architecture + Planning selected five TODs across the country, each with a slightly different approach to development and parking — Englewood, Colo.; Wilshire/Vermont in Los Angeles; Fruitvale Transit Village in Oakland, Calif.; Redmond, Wash.; and Rhode Island Row in Washington, D.C. The research team counted the number of people entering and exiting the TOD buildings and conducted brief intercept surveys of a sample of them. The team also counted parking inventory and occupancy.

    The study found that all five TODs generated fewer vehicle trips than standard guidelines estimate, and used less parking than many regulations require for similar land uses. Most of the TODs included in this study also built less parking than recommended by engineering guides, yet even this reduced amount of parking was not used to capacity. The ratio of demand to actual supply was between 58 percent and 84 percent. Fewer vehicle trips is one likely reason why parking occupancy rates were lower than expected. Another possible reason is that standard engineering guidelines do not fully account for other travel modes that are available and actively encouraged at TODs.

    Empty Spaces: Real parking needs at five TODs is available at

    Mary Smith, P.E., is senior vice president and director of Parking Consulting for Walker Parking Consultants. She may be reached at