Depending on your point of view, you either love parking or you hate it. Municipalities love the revenue it generates. Drivers hate searching for it.
A new solution offers a way to please both sides. CloudParc combines machine vision and artificial intelligence
(AI) to increase parking revenues and reduce search times. The best part? No parking meters needed.
Evaluations are currently underway at the Miami Parking Authority and the Charleston Urban Renewal Authority.
Where today’s tech falls short
1. Traditional parking meters are expensive to install and maintain. They’re also vastly underpowered compared to what is now possible with parking technology: They can’t report whether a parking spot is occupied nor can they adjust the meter price according to demand.
2. Car sensors embedded in the road are better than meters, but they’re expensive to install and upgrade and what the technology can do is limited. It can only tell you if the space is occupied. It does not link up with payment options, and it cannot help identify the parked vehicle.
3. "Pay-and-Display" meters can cover many parking spots but are inconvenient for drivers. They cannot report when a spot is vacant, nor do they offer remote notification and/or refills.
4. Mobile, pay-by-phone apps help drivers find empty spots and pay fees, but still depend on a physical parking meter.
What’s needed is an approach that automatically reports space availability, electronically ID’s the registered vehicle as it parks and starts charging for time in the space. There’s no need for meters. This is the ultimate in parking convenience for drivers, like automated toll collection, only for parking.
Designing the solution
1. Identifying a parking spot
High-resolution, connected cameras are widely used for security monitoring. Installed on electric poles, they can affordably monitor large areas. Applying artificial intelligence to the images allows the system to learn to recognize parking spots on a street. They can also report which spots are occupied and for how long.
2. Which spot is vacant?
Feeds from the cameras are sent to the cloud for analysis. AI recognizes which spots are vacant and sends this information to mobile apps that inform drivers about available parking spots.
3. How long was the car parked?
License plate recognition software is used to accurately identify the vehicle parked in a particular spot. The camera also tracks how long the vehicle is parked in that spot. If the driver is not registered with the application, the car’s license plate details are used to bill the driver.
4. Dynamic pricing
CloudParc can continuously track occupancy patterns for every block, generating a heat map. This data can help inform intelligent pricing decisions. Pricing can be easily changed in the cloud depending on the time, day and other factors such as local events in order to maximize revenue and reduce congestion. It can also give valuable data to the local government on where new parking spots are needed.
The company offers basic services free of charge, recouping its costs through revenue-sharing agreements with the parking entity — public or private.
Deepak Puri, a former Oracle and VMware executive, is the founder of SkilledAnalysts.com, a San Francisco-based IoT knowledge firm.