Researchers at the University of South Australia (UniSA) are fueling the waste management game by deploying artificial intelligence (AI) in bins to monitor public garbage collections.
The technology will be used to predict which places produce more waste so authorities can plan how often public bins need to be emptied.
UniSA PhD candidate Sabbir Ahmed is designing a deep learning model that uses algorithms to analyze data from smart bin sensors.
“Sensors in the public smart bins can give us a lot of information about how busy specific locations are, what kind of waste is being thrown away and even how much methane gas is being produced from food waste in bins,” Ahmed said.
By feeding all the data into a neural network model, the design can predict which public bin locations fill up quickly and which are barely used.
The insights can help municipalities optimize their waste management services, plan dumpster cleanups, and even move dumpsters to areas of higher demand.
Ahmed is partnering with Wyndham Council in Victoria on a pilot project that uses smart bin data to develop an AI model.
The research is published in the International Journal of Environmental Research and Public Health.
Co-author of the paper, UniSA lecturer Dr. Sameera Mubarak said waste management is a growing concern worldwide.
“Many urban areas are struggling with an increase in waste due to rapid population growth, and waste services are becoming increasingly difficult for local governments to manage,” she said.
The researchers are examining sensor data from public bins, routing paths and collection locations to develop their AI model.
The sensors can detect different types of waste, including solid, organic, industrial or chemical, medical and recycling.
AI can quickly predict patterns of waste generation in public spaces, such as identifying busy days and upcoming events that could lead to more waste.
This allows you to better plan waste collection and prevent overflowing bins.