AI Governance in the Smart City: A case study of garbage truck mounted machine vision for roadside maintenance

This case study involved adopting a 5G mobile IoT solution for automatic roadside asset condition monitoring (5G AI solution), a collaboration between Swinburne University of Technology and Brimbank City Council, funded by the Australian 5G Innovation Initiative. This smart management solution analyses camera data from camera-mounted waste collection trucks to automate the identification of roadside assets requiring maintenance and enhance asset management efficiency. As with all data-driven and AI technologies, there are risks and governance requirements. The 5G AI solution can gather sensitive information, potentially leading to 'scope creep' (extension of use beyond the initial goals) or data misuse. It needs investment in infrastructure, management and maintenance, but doing so will present substantial long-term efficiencies and benefits. A key finding of the project is that AI ethics principles must be adapted and translated to improve governance within organisational settings such as local government. This report describes the steps that should be taken to align governance actions with ethical AI principles.