Session 04: Automating Maintenance in the Smart City: AI Governance for Mobile Machine Vision
SESSION 04: MOBILITIES SYSTEMS AND DEPLOYMENT
This presentation delivered by Dr Yong-Bin Kang was recorded at the 2024 ADM+S Symposium.
Automating Maintenance in the Smart City: AI Governance for Mobile Machine Vision
Dr Yong-Bin Kang (Swinburne University of Technology), Professor Anthony McCoske (Swinburne University of Technology) & Dr Milovan Savic (Swinburne University of Technology)
As smart city initiatives proliferate globally, artificial intelligence (AI) systems are increasingly deployed to provide automated services. However, while momentum is building around AI ethics and regulation, there is uncertainty about how these principles and frameworks are to be transformed into AI governance to guide use cases. This paper explores our collaborative research with Brimbank City Council, and what it takes to establish on-the-ground AI governance processes for mobile machine vision systems. Brimbank Council has deployed a 5G AI solution that utilises cameras mounted on garbage trucks to capture road asset information and detect maintenance needs as they move about the neighbourhood.
This solution empowers the council to manage assets more efficiently, cuts costs, and leverages data-driven decisions to benefit the community. However, implementing responsible AI and building AI governance in practice requires a certain kind of ‘edge work’, an active process of translation and alignment with existing institutional practices and policy frameworks. In emerging AI governance literature, this is understood to involve important scaffolding work necessary to transform principles in to practices and actions. Our co-designed AI governance framework offers lessons and insights for AI governance that other local governments and communities can adopt. Project report available at: https://apo.org.au/node/323811