/ smart parking

KickView Smart Parking Application

The automotive industry is moving from an "ownership" model to a "services" model, with services focused in two areas: (1) The service of transportation, and (2) services delivered and sold inside the vehicle. In addition, vehicles are becoming more connected to the internet with technologies like 5G driving adoption even faster. Drivers have more options to consume transportation related services. However, on a daily basis, too many people are dealing with the hassles of finding appropriate parking. Therefore, one service that can greatly benefit many people is a streamlined parking solution.

Figure 1. KickView Smart Parking application with multi-camera association and fusion. Cameras cover front half of parking lot. Third camera will be installed later for fusion of far section.

According to a study by INRIX in 2017, Americans spend an average of 17 hours per year searching for parking, resulting in a cost estimate of $345 per driver in wasted time, fuel and emissions. BMW claims that up to 30% of traffic congestion in major cities is caused by people looking for parking. There seems to be no dispute among drivers and city operators that a more efficient method of finding and utilizing parking is needed. Car manufacturers are already preparing their fleets with key services that include an eye to the future of parking. BMW acquired ParkMobile with the goal of becoming the world's leading provider of digital parking solutions. ParkMobile provides an app that allows users to find and pay for parking from their phone. Parking.com provides event-based parking and Google Maps on the iPhone allows you to search for parking at the destination you've chosen and add that as part of your route. While several companies are building the backend infrastructure for offering location based services, there remains a gap in the technology for streamlining the real-time sensing required to get precise parking slot usage, availability and surounding analytics. Current large scale applications lack detail needed once you get to an actual parking lot (e.g., which spots are currently open, and best route to an open parking spot). Also, important analytics such as: (1) how long do drivers spend searching for a slot, (2) parking occupancy over time, (3) turnover, (4) traffic counts and speeds in parking area, and (5) incident detection (e.g., how safe is this lot), are important metrics for drivers, parking lot and smart city operators.  

Figure 2. Real-time AI smart parking application with geospatial tracking and precise parking slot occupany determination.

There exist proposed solutions that use wireless sensors, such as sensors embeded in the ground at each parking slot, or sensors mounted on a wall/pole at each slot. While these solutions can be effective, they are expensive and require significant installation complexity, cost and risk. For example, installing sensors into the pavement will require a complete reinstall when the parking lot needs resurfaced or repaired. In regions with severe winter weather, this can be very inconvenient.

The KickView smart parking solution utilizes passive video camera sensors and real-time intelligent video analytics. Overall, this solution is ideal from a cost, complexity and integration perspective. The KickView intelligent video solution provides you with valuable insights that improve efficiency. Benefits of KickView's smart parking solution include:

  1. Constant monitoring of all vehicles and the surounding area - providing situational awareness of operations, incidents and anomalous behaviour.
  2. Accurate counting (vehicles, bikes, pedestrians, etc.) and precise estimation of which parking slots are occupied or vacant.
  3. Parking spot/lot usage over time, peak traffic times, average time spent searching for a parking spot, and more.
  4. Simple integration with large-scale infrastructures or mobile apps.
  5. Vehicle detection, license plate detection, geospatial tracking from entry to exit.
  6. Sensor fusion across multuple video cameras for complete situational awareness from a single map-view.
  7. Incident detection and automatic alerting.
  8. And more...

Figure 3. An example of a processing pipeline in the KickView kvSonata framework.

KickView's smart parking application is built using the KickView kvSonata framework. kvSonata removes the complexity of building a sensor data pipelining architecture from scratch. If your operational needs change, you will be able to easily define new processing workflows and configurations without losing valuable time worrying about redesign and low-level architecture. In addition, kvSonata simplifies the task of customizing any real-time AI models so that processing capabilities adapt with your operational needs.

Figure 4. KickView workflow with 3rd party license plate detection. KickView algorithms used for association and geospatial tracking.

Contact us at KickView to learn more about how we can help you streamline your video analytics solutions. KickView is a world leader in real-time AI detection and geospatial tracking using video. You can also Contact us to learn more about licensing advanced analytics capabilities for your applications.  

KickView

KickView

KickView provides real-time AI solutions for extracting actionable information from sensor data.

Read More