By Christina Onesirosan Martinez
I bravely volunteered to write this post on the role of artificial intelligence (AI) in the parking industry and my initial research took me to the first floor of our office—deep into the darkest corner of the research team’s lab. They spend hour after hour looking at the best way to answer a question on everyone’s lips at last once a day: Where can I park?
The size and diversity of the data available today allows researchers to build mathematical models to not only accurately find parking and predict space availability, but also start enabling fully autonomous parking.
The team has spent years developing predictive algorithms to give drivers an indication of availability at a parking lot before they arrive. For example, a parking lot with sensors installed is able to predict availability to 95 percent accuracy.
The more I listened, the more I learned about how they use machine learning (ML) and deep learning (DL), branches of AI, extensively for everything from ingesting and processing billions of data points per day to reading parking signs to enabling autonomous parking.
These data points include, for example, data that is already used to create real-time traffic information. The same data is used from the same systems but instead of working out how quickly vehicles are moving, they work out how they are moving. Are they circling an area to find a space? This data is then used to figure out if there is a space available or not. This is just one example of how machine learning can be leveraged by the parking industry to benefit drivers.
One of the downsides is the lack of connectivity as well as data sharing in the parking industry. Consequently, they have to rely on a wide range of additional data sources that can be used for predictions. This can include data on local conferences, sporting events, and concerts. This means the team can work from billions of data points daily to process, learn, and infer where parking will be. It’s not 100 percent precise but for now, it will suffice!
Christina Onesirosan Martinez is head of marketing and operations for Parkopedia.