When you look across all of the public parking spaces in the urban areas, a very small percentage are being monitored for whether the spot is open or occupied. There could be surveillance video cameras mounted on poles that can do image recognition, or perhaps there are sensors embedded in the pavement that can detect occupancy. [e.g., Bosch, earlier post] Both of those sources provide very good high quality data. The issue is a large cost onus on the infrastructure owner, the parking lot owner to install that hardware.

Perhaps we as an automaker can use our vehicles as mobile probes. The sensors that are on those vehicles can be used to detect open parking spaces. We can achieve the end game of a deeper or richer data environment. However, it has to fit together properly.

In a practical sense, parking space occupancy is very time-sensitive. A space can be open one second and literally not the next second. One of the things we are working on is adding in a probability factor. If a car drives by, and transmits data on an open space up to the cloud, will it be open 5 or 10 minutes later? What’s the probability that the spot is truly open? That’s part of the smarts that need to be built into this. There is a lot of data out there from which models can be built. Spatial, temporal … to make it truly valuable, we need to build it into the system.

—Dave McCreadie

The urban parking problem. Congestion is one of the most prevalent transport problems in large urban agglomerations, usually above a threshold of about 1 million inhabitants, notes Dr. Jean-Paul Rodrigue at Hofstra University. Dr. Rodrigue sits on the international editorial board of the Journal of Transport Geography and the Cahiers Scientifiques du Transport, and acts as the Van Horne Researcher in Transportation and Logistics at the university.

Increasing use of automobiles has correspondingly increased the demand for transport infrastructures; however, the supply of infrastructures has often not been able to keep up with the growth of mobility.

Since vehicles are parked for the majority of the time, increased automobile use has expanded the demand for parking space, which in turn has created space consumption problems particularly in central areas; the spatial imprint of parked vehicles is significant, Dr. Rodrigue says.

Congestion and parking are interrelated because looking for a parking space (“cruising”) creates additional delays. Rodrigue notes that in central areas of large cities, cruising may account for more than 10% of the local circulation as drivers can spend 20 minutes looking for a parking spot.

In his oft-cited book on parking and policy (“The High Cost of Free Parking”), Dr. Donald Shoup, Distinguished Professor of Urban Planning at UCLA, noted that 16 studies conducted between 1927 and 2001 found that, on average, 30% of the cars specifically in congested traffic were cruising.

New studies continue to find that many drivers cruise for curb parking like hawks looking for prey. For example, when researchers interview drivers who were stopped at traffic signals in New York City, they found that 28 percent of the drivers on one street in Manhattan and 45 percent on a street in Brooklyn were cruising for curb parking.

In another study, observers found the average time to find a curb space on 15 blocks in the Upper West Side of Manhattan was 3.1 minutes and the average cruising distance was 0.37 miles. These findings were used to estimate that cruising for underpriced parking on these 15 blocks alone create about 366,000 excess vehicle miles of travel and produces 325 tons of CO₂per year.

—The High Cost of Free Parking

Shoup advocates performance parking—i.e. variably setting prices to produce one or two open curb spaces every city block—as one solution for the parking/congestion issue. To set the right price, however, the city managers need to be able to observe occupancy; Dr. Shoup notes that occupancy sensors are one potential contributor to a viable solution.

A 2013 survey by parking provider APCOA found that the search for parking space in Germany on average requires 10 minutes, in Italy, 15 minutes. According to the study, a parking search, ranging over 4.5 km (2.8 miles) on average, costs car owners €1.35 (US$1.54), with an environmental impact of 1.3 kg of CO₂ emissions.

We knew traffic resulting from the search for parking space is being underestimated, but this drastic result really surprised us. If car drivers would directly steer for the next car park, they would save an enormous amount of time and money.

—Ralf Bender, Chief Executive Officer of APCOA Parking Group

APCOA has been advocating improved parking traffic guidance systems in 12 European countries, and made sure all its car parks were programmed into the traffic navigation systems of the major providers.

Ford’s initial smart parking projects. The initial smart parking projects that are part of the Smart Mobility Plan include:

• The Parking Spotter experiment, conducted with Georgia Tech, leverages driver-assist sensors that most Ford vehicles already have, including sonar and radar. The sensors search for open parking spaces while the driver looks for spots around the city, and share the information with a cloud database other drivers can access. The system makes it easier for a driver to locate an open spot, reserve it and navigate to the space.

  We have been in experimentation mode for the last year to see whether or not our vehicles can do the sensing required and then send that data off-board to the cloud. That’s the scope of what we are trying to do. This very much resides in the IT space. On the one hand, we have to have the correct sensors on the car, but then it becomes data management. Aggregating your data with similar data coming form other vehicles, maintaining a map of open spaces.

  —Dave McCreadie

• The Remote Repositioning experiment, which uses on-vehicle cameras and live streaming to allow drivers to control vehicles from miles away.

• The Painless Parking project, which helps drivers pay for parking meters through a mobile phone.

Ford is not trying to get into the business of creating parking apps, McCreadie said; rather, Ford wants to make the apps better by providing more data to support them. Ford did write the software that did the analysis of the sensor data. However, once you get beyond the software, McCreadie added, the “other critical thing is the power of numbers.”

If there is only a handful of Ford vehicles driving around major metropolitan areas, the benefit is not as great. What we need are thousands of vehicles that are doing that. The best way to do that is to start to engage other automakers. Frankly, they have the same types of sensors in the front and rear fascia. They generally have similar hardware. The question now is can we partner with them? Bring them in the paradigm to collaborate together?

—Dave McCreadie

Such an effort will entail standardization on communications and data formatting, although Ford is not yet at the stage of promoting such efforts.

We are still experimenting. It’s clear to everyone that parking is an issue. It would be a huge benefit to customer convenience and perhaps even cost and the environment if we could all do a quicker, better job of parking our cars. Stay tuned for more from Ford; there is a huge carrot out there.

—Dave McCreadie

Resources

• Douglas A. Thornton, Keith Redmill, Benjamin Coifman (2014) “Automated parking surveys from a LIDAR equipped vehicle,” Transportation Research Part C: Emerging Technologies, Volume 39 Pages 23-35 doi: 10.1016/j.trc.2013.11.014

• Christopher McCahill , Norman Garrick (2014) “Parking Supply and Urban Impacts,” in Stephen Ison, Corinne Mulley (ed.) Parking Issues and Policies (Transport and Sustainability, Volume 5) Emerald Group Publishing Limited, pp.33 - 55

• Shoup, Donald. The High Cost of Free Parking. N.p.: Natl Book Network, 2011. Print.