A New Parking App, No New Sensors Needed

Big-data company INRIX is increasingly reliant on location data crowdsourced from drivers using its navigation app for visibility into traffic—and now, parking availability.
Published: June 10, 2015

Cities have long been installing sensors along roadways—embedding them in streets or mounting them overhead—in order to better understand everything from traffic flow to parking availability. And as in-car navigation systems and smartphones have proliferated, INRIX, a company headquartered near Seattle, Wash., has for the past 10 years built a suite of products that leverage sensor data to provide actionable data for drivers and cities alike.

For example, the INRIX Traffic application, which is accessible via a smartphone app and through in-car navigation systems of participating vehicles (including those from Audi, Lexus, Volkswagen, BMW, Tesla and Toyota) helps drivers find the fastest routes from point A to B in cities throughout more than 30 countries.

BMW’s parking app

Last week, INRIX announced its newest product, INRIX On-Street Parking, an application that directs drivers to streets with the most abundant available parking spots, pricing information for those meters, and location and pricing information for nearby off-street pay-to-park options. BMW will be the first automaker to introduce the product, as part of a new navigation system it is debuting in its 2016 models.

But over time, says INRIX director Jim Bak, the firm is becoming less dependent on purpose-built traffic sensors. More and more, the company is able to leverage the location data transmitted by the smartphones and in-vehicle navigation systems running its application. As this “crowdsourced” data set grows, he explains, it is opening their services up to cities that lack traffic sensor networks.

“The challenge for cities is that to scale sensor networks, they have to invest and put them everywhere,” Bak says, noting that San Francisco decided to stop expanding its SFPark program, which began piloting in 2006, as the batteries that power the sensors started to reach the end of their lifecycles. (The city’s SFPark smartphone app still shows where parking meters and garages are located, but no longer provides availability data.)

INRIX is able to power its on-street parking application by collecting data from cities regarding the use of Internet-connected parking meters—to estimate occupancy, it looks at when transactions begin and when the meter’s credit is due to run out—combined with anonymized location data from INRIX app-running phones, as well as from the cars with INRIX’s Traffic app (transmitted through the embedded cellular modem inside those vehicles).

Speed is an important behavioral cue, Bak notes, since people who are looking for parrking tend to drive more slowly down a block, scanning for an open spot, than those just passing through an area.

He says INRIX’s algorithms crunch the collected data to estimate parking availability.

“We fill in the gaps” in places where sensors are not monitoring traffic or parking-spot occupancy, Bak says. “Our point of view is that you can provide services for drivers, and get the data that powers those services, in a way where you don’t have a scalability problem because you can collect data where the cars go. Instead of making the road the sensor, why not make the car the sensor?”

INRIX is rolling its on-street parking availability app to carmakers and other companies that want to use the data for their own navigation-system applications. The first cities in which INRIX has sufficient data to offer the service are Amsterdam, Cologne, Copenhagen, San Francisco, Seattle and Vancouver, though it plans to begin offering the service to other congested cities as well, such as Paris, London and Chicago. According to Bak, his company has tested results from its parking-spot availability algorithm in San Francisco, and has found the algorithm’s results to be 80 percent accurate compared with those derived using data from the still-active occupancy sensors throughout the city.

A study conducted by the planning department at UCLA estimated that at some parts of the day, up to a third of drivers in urban corridors are not actually going anywhere—they’re circling blocks, looking for parking. Helping drivers to quickly find parking spots not only reduces congestion and carbon emissions, but also benefits local businesses by making it easier for their customers to park. Cities can also benefit from the traffic and parking information that INRIX provides, since they can use it to understand how well or poorly parking meters are being utilized. Cities are increasingly using meters that can change fees dynamically throughout the day, with prices increasing during times of highest demand.

INRIX can provide cities with traffic and parking data that is updated every two minutes, Bak says, at a quarter of the cost of installing and maintaining sensors along roadways.

The big-data company is now developing an application for which it is working with carmakers to collect information from a range of in-car sensors—including those that measure outside temperatures or track the operation of a vehicle’s braking system or windshield wipers—to generate reports on weather correlated to the route that a particular driver is taking, based on the destination he or she has entered into the navigation system. INRIX would also share this information with the U.S. Departments of Transportation in order to help them dispatch resources (such as plows or road salt trucks) where they appear to be needed most, based on the sensor data and the speed of cars on the roads.