Google maps predicts parking difficulty using machine learning

Sustainability problem: Urban mobility

Drivers usually circle the blocks around their destination to find parking, and over 30% of traffic in cities is caused by these cars looking for a parking spaceThe problem is that there is almost no real-time information about available parking spots. Even in cities with smart parking meters that make this information publicly available, this data doesn’t account for those who park illegally, park with a permit, or depart early from still-paid meters.

Technology solution: Machine learning prediction

  • Earlier this year, Google Maps started showing an icon with predicted parking availability when displaying driving directions. This feature is available in 25 cities across the US.
  • This new feature works using a combination of crowdsourcing and machine learning algorithms that predict parking difficulty in a certain destination, based on anonymous aggregated information from users who opt to share their data on Google Maps or Waze.
  • “Parking difficulty” is estimated by identifying users that circled around a destination instead of arriving right away to a place. The more circling, more difficult the parking in that area must be. Using this information, the machine learning model assigns a descriptive prediction of parking availability to display to the user, like “Easy” or “Limited parking”.
  • In a pre-launch experiment, Google researchers saw a significant increase in clicks on the transit travel mode button, indicating that users with additional knowledge of parking difficulty were more likely to consider public transit rather than driving.

Organizational stakeholders

  • City Officials
  • Department of Transportation
  • Parking lot owners
  • Local communities

Implementation steps (for a more integrated solution)

  1. Connect with City Officials and Department of Transportation to incorporate traffic data sources into the model
  2. Connect with parking lot owners to incorporate real-time information of the parking availability on their garages
  3. Scale the solution to other cities.


Comment: Just when you thought it couldn’t get better… HomeBioGas 2.0

“One great feature of this solution is that it is can be assembled by anyone. It comes with a DIY kit and easy instructions, so there is no need for a professional technician and you get the system ready in a few hours.”


2 thoughts on “Google maps predicts parking difficulty using machine learning

  1. I imagine this application can be extended to other supply/demand service scenarios such as assessing wait times for entering events (i.e. concerts or baseball games) or restaurant seating availability. Certainly this will lead to less demand for crowded services, and thus an influx of supply! “Nobody goes there anymore – it’s too crowded.”


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s