Minnovation has created a car park monitoring system for the City of Banyule that utilizes advanced hardware, software and deep learning technology to accurately recognise vehicles entering and exiting parking lots. This system provides a reliable and efficient solution for tracking cars in car parks, thus improving the safety and security of drivers and their vehicles.

The City of Banyule is a local government area in Victoria, Australia in the north-eastern suburbs of Melbourne. It was created under the Local Government Act 1989 and established in 1994 as an amalgamation of former councils. It has an area of 63 square kilometres (24.3 sq mi) and lies between 7 and 21 km from central Melbourne. Wikipedia

The Yarra River runs along the City’s southern border while its western border is defined by Darebin Creek.

Problem

The city wanted to find some innovative ways of monitoring the utilisation of three of its public carparks in geographically distributed areas.

The solution needed to be lower cost than deploying in-ground sensors and had to be able to provide occupancy/vacancy numbers in real time.

The solution did not have to be enforceable and was intended to provide indication to the public, rather than detecting overstays. The solution was also not intended to detect the utilisation of individual bays, the city had decided that it was more important to display overall occupancy/vacancy statistics to the public.

Solution

The solution utilised a Minnovation supplied AlphaX Vision™ hybrid solution.

For the initial installation, a total of 5 seperate cameras were installed. Cameras were installed at the entrance and exit of two (2) of the carparks, and at one (1) car park, a single camera was used to monitor the single entrance.

The cameras utilised deep learning algorithms to detect vehicles as they entered and exited the carpark. Each of these values were tracked individually and were combined to calculate an overall occupancy using the AlphaX Cloud virtual channel feature.

As the algorithms are running on the edge processors at the camera, personally identifiably information is not collected or stored. Instead only the metadata of the vehicle tracking is sent.

The cameras are connected to a solar power system and mounted on a 5 meter post to provide an adequate height to detect the vehicles. The battery and battery management system are able to run the system for a full 7 days without any sunlight providing ample power in the event of bad weather.

The AlphaX Cloud platform allows the City to perform analytics of the data as it is received. Occupancy is updated on public facing dashboards for public to access from a unique URL. Real time alerts are also in place to help monitor and manage parking bottlenecks.

The City has established auto generating reports for their data so they can monitor utilisation throughout the month. These reports are automatically sent to planning teams so they feed future planning needs.

A smart park application has been created for previous projects and the parking data is fed into this application so that the public can access car park occupancy numbers.

Digital signage marks the final piece of this project, providing the public with instant vacancy numbers upon their entry to the public car spaces.

Outcomes

The project helped the city save almost 60% of the cost for implementing an in-ground solution, representing a significant cost saving when compared with projects using a alternative technologies.

The accuracy of the system has been measured to be in the 93-95% range when compared with manual counting. The margin of error typically occurs when a larger vehicle travels in between a smaller vehicle and the view of the camera. A mounting change would see this margin of error decreased.

The systems insights are shared with a number of teams and third parties within the council asset management initiatives and are used to feed data to 3rd party applications using the AlphaX Cloud platform.

This project was part of the Cities asset management initiatives 2022 which won the People Choice Award and was highly commended in the Data and Insights category at the MAV 2022 Awards for Excellence.