Autonomous Vehicles and the Hidden Parking Problem: Where Will We Park Self-Driving Cars?
Autonomous vehicles are coming. That’s led many people to declare that parking will become obsolete.
However, the truth is quite a bit more complicated. Autonomous vehicles will still require places to park. Moreover, non-autonomous vehicles aren’t going to disappear overnight. These truths together reveal a clear need to devise parking solutions that work for both autonomous and non-autonomous vehicles.
The parking challenge posed by autonomous vehicles
In recent years, a certain vision of autonomous vehicles has emerged. In this vision, a fleet of autonomous vehicles are constantly circulating the streets, providing on-demand rides to people who need them. Without the need to park, the whole fleet will be optimized for efficiency. No autonomous car will drive further than is absolutely necessary.
Autonomous vehicles will still require places to park. Moreover, non-autonomous vehicles aren’t going to disappear overnight.
But this model has several logistical flaws. The demand for rides in autonomous vehicles will fluctuate throughout the day. There will be times when there is a very low demand for rides — say, 4:00 in the morning. Conversely, other times of the day will see demand spike.
If autonomous vehicles are to serve riders efficiently during peak periods, then a large fleet will be necessary. Unfortunately, those cars aren’t just going to disappear during low demand periods. If the fleet is constantly circulating during the off-hours, autonomous vehicles are not increasing overall efficiency–they’re just contributing to traffic congestion and consuming energy for no reason.
The bottom line: Autonomous vehicles will need parking. But figuring out where that parking should be presents its own challenges.
Considerations for parking autonomous vehicles
Figuring out where to park all of these autonomous cars poses numerous challenges.
First, it’s likely that many (or even most) autonomous cars will be electric vehicles. That means parking spaces must also include charging stations. And that means more investment in infrastructure.
Downtown real estate is expensive, so creating autonomous vehicle parking/charging structures there may not be economically feasible. But that creates additional complications — especially for electric vehicles.
Think about this scenario: An autonomous vehicle has about 15% of its charge left. The charging station is located in an outlying area of the city. There is no rider downtown right now but one is predicted in the next 10 minutes. So, should the car wait somewhere in the hope that the trip will be short enough to make sure it returns to the hub with 2–5% charge left? Or does it need to go charge right now?
The answer to this dilemma is by no means obvious. Yet for autonomous vehicles to work on a large scale, we need to figure it all out for the fleet to operate effectively (let alone efficiently).
Additionally, we must figure out a parking system that is fair for both autonomous vehicle riders and traditional drivers. Even once autonomous vehicles enter the market, non-autonomous vehicles will continue to be on the road for some time.
This reality significantly impacts how we tackle the autonomous car parking problem. Although one potential solution is to create designated parking spaces for autonomous vehicles, such a solution is likely to cause social backlash. After all, early adapters of autonomous vehicles are likely to be wealthier than most. So if we set aside autonomous vehicle-only parking spots, it’s essentially giving away valuable parking spaces to well-off people. That’s unlikely to be a popular position.
We must figure out a parking system that is fair for both autonomous vehicle riders and traditional drivers.
Finally, how do we make sure that autonomous vehicles are properly charged for parking and curbside usage? The city cannot install and maintain reliable sensors across its entire road network while there is no driver to be held responsible for parking violations. There are over 3 million parking spaces in New York, most of which are non-metered and governed by alternate side parking regulations. How do we make sure that self-driving vehicles abide by these complex rules and, if they are ticketed, how do we make sure that they move from their location at the earliest available time?
In short, figuring out where and how to park autonomous vehicles is a problem that requires logistical problem-solving as well as a deep understanding of social context.
How data can address the problem
It’s important to note that this problem is by no means intractable. However, it will require massive amounts of data. In order to design parking for the autonomous vehicle future, need to understand how drivers are currently parking. Data collection is step number one.
Parkofon is collecting data on parking from the driver’s perspective. We’re learning critical information about where people park, when they park, and how long they stay parked. This data is critical for figuring out how autonomous and non-autonomous vehicles can share parking spaces in the future.
As we move towards autonomous vehicles, we need parking data from a wide variety of sources. Once we understand how people park, we can help cars park without people.
There’s other promising work being done in this space. Working in conjunction with the University of Barcelona, the European Parking Association (EPA) is conducting a study to learn more about parking availability in 21 European countries — covering 41 million parking spaces in total.
Increasingly, drones are also being used to collect parking data. Drones can offer an aerial perspective on parking availability and usage patterns.
As we move towards autonomous vehicles, we need parking data from a wide variety of sources. Once we understand how people park, we can help cars park without people.
