Making Science

Nokia Scene Analytics AI tech powers Melbourne waste-dumping trial

David Murphy

Nokia and the city of Melbourne have released details of a trial conducted using Nokia Scene Analytics Artificial Intelligence (AI) technology to develop a deeper understanding of waste disposal behaviour. The aim of the trial is to enable the city to tackle the issue of waste dumping more efficiently and keep laneways – the busy and narrow city streets and pedestrian areas – cleaner, safe and free of garbage.

To decrease the frequency of waste contractor visits to busy areas, the city of Melbourne has offered local residents and businesses subscription-based access to large-capacity compactor facilities. With the compactor in place, the city council then wanted to understand how the service was being used and how to mitigate illegal waste dumping, which can quickly create safety and hygiene issues in the area.

Under its ‘emerging technology testbed’ initiative, the city worked with Nokia to leverage an existing network of installed cameras as Internet of Things (IoT) sensors to monitor one of the compactors. The Nokia Scene Analytics solution employed an AI-powered algorithm to filter and collate data from the cameras, while also bringing in other data sources, such as operational data on the compactor itself, to create real-time alerts and produce reports. 

The trial allowed for real-time monitoring and detection of activity in the vicinity of the compactor using a virtual tripwire. Object detection and object counting was used to identify and count items, to show how the compactor was impacted by items incorrectly placed within it, while also identifying potentially dangerous items. Anomaly detection identified unusual movements, such as illegal waste dumping during the night, while face and license plate blurring maintained individual privacy during the trial.

Using these reports, City of Melbourne can better understand the correlation between illegal waste-dumping activities and compactor downtime, to keep maintenance teams better informed and minimize issues. It also allows them to swiftly address waste dumping activities before they become a hazard, viewing locations in real-time to observe any obstructions to service vehicle access, and adapting their schedule to reduce unnecessary visits.

Nokia said that initial trial results demonstrate that Scene Analytics can support the city’s objectives for better, safer citizen experiences, while simultaneously lowering maintenance and down-time costs for waste management services.

“This is a great example of using new technology to help remove illegal waste more quickly, make our city cleaner and protect the environment,” said Melbourne Lord Mayor Sally Capp. “Our partnership with Nokia is another way we are gathering data to make Melbourne a safer, smarter and more sustainable city. This innovative project will help to avoid hazards and make our streets even cleaner by allowing our waste services to better understand behaviour trends related to the illegal and dangerous dumping of waste.”