The City of Canterbury-Bankstown has invited organisations to a market sounding session during the week of 23 October to discuss improvements of its AI recycling contamination model.
In 2020, the south-western Sydney council became the first in Australia to launch an AI model that automatically detects waste contamination in recycling bins.
The model was developed by the council using Microsoft Power Apps to help solve the issue of residents placing the wrong items in recycling bins.
Prior to using the AI model, the City of Canterbury-Bankstown's recycling waste contamination rate was 25 per cent, compared to the state average of 10 per cent.
The AI model analyses data captured by GPS and Global Navigation Satellite System (GNSS) cameras installed on the council's fleet of 19 waste trucks, which scan recycling bins as they are tipped into trucks.
The model then provides reports of contamination types, locations and heat maps, which are used to inform the council's recycling education programs in specific areas.
Part of the council's $2.1 million technology transformation program, the AI model won the Outstanding Council Project category at the 2023 National Waste Conference.
The City of Canterbury-Bankstown's sounding session will offer confidential one-on-one discussions for organisations and software developers during the week starting 23 October.
The one-hour sessions will allow for discussion with the council's waste operations and IT teams on the needs and possibilities of a future AI model
The council will use information from the sessions to compare AI and machine learning models with its existing AI model and inform the creation of an RFQ for a possible future tender.
-1.jpg&w=100&c=1&s=0)
.png&w=100&c=1&s=0)
