The Victorian government's ambitious plan to enhance road safety through AI-powered cameras has hit a major snag. A scathing audit by the Victorian Auditor-General's Office (VAGO) has revealed a multitude of issues with the Distracted Driver and Seatbelt Detection Program, raising serious questions about its effectiveness and the integrity of its implementation.
The audit, which examined 308 camera sites, 160,754 camera hours, and 188,493 infringements, found that the program's impact on road safety is unclear. The report stated that the number of road deaths in Victoria has increased since the program's inception, contradicting the government's claims of improved safety. This alarming trend is particularly concerning given the program's substantial cost and the ambitious goals set by the Victorian Road Safety Strategy 2021-2030.
One of the most significant shortcomings identified was the department's inability to measure the program's success. The Department of Justice and Community Safety, responsible for the program, lacks the necessary data to determine if the program is reducing accidents, injuries, and fatalities. This is largely due to poor record-keeping, inconsistent reporting, and a lack of baseline data for comparison.
The audit also uncovered privacy breaches and data accuracy issues. The use of self-reporting by Jenoptik, the company providing the AI software, raises concerns about data integrity. The department's reliance on this system means they are unable to fully manage the risk of inaccurate data, which could have serious implications for road safety decision-making.
Furthermore, the process for selecting camera sites is flawed. Victoria Police, tasked with identifying these sites, does not consistently record its decisions, making it difficult to demonstrate that camera locations are based on sound evidence. This lack of transparency and evidence-based decision-making undermines the program's credibility.
The audit's findings are particularly damning given the government's recent plans to invest $28.3 million in next-generation cameras capable of detecting a wider range of offenses, including speeding and red-light violations. This investment, intended to further enhance road safety, is now in question due to the program's current shortcomings.
The VAGO's report recommends significant improvements in data collection, reporting, and governance. The department must address the lack of baseline data, improve record-keeping, and ensure the integrity of the AI software and its data. These changes are crucial to restoring public confidence in the program and ensuring that future investments in road safety technology are effective and well-managed.
In conclusion, the Victorian government's AI-powered road safety program is facing a critical moment. The audit's findings highlight the need for urgent reform to address the program's shortcomings and ensure that road safety initiatives are based on robust evidence and effective governance. The future of road safety in Victoria depends on these necessary improvements.
