How AI-Powered Video Analytics Is Improving Transport Safety and Infrastructure Management

How AI-Powered Video Analytics Is Improving Transport Safety and Infrastructure Management

Authorities are using near-real time generated AI-powered video intelligence to better understand risk, improve safety outcomes, and support infrastructure management.

In Western Australia, initiatives by Main Roads WA, local governments, and the Road Safety Commission (RSC) have included the deployment of SensorMine which is a single IoT platform delivering near-miss detection, traffic analytics and road-safety insights. AI transforms camera feeds into proactive infrastructure decisions at scale.

SensorMine was developed through a partnership between LAB3 and strategic technology partner Altra. Altra brings expertise in AI, IoT, sensing and operational technology, while LAB3 contributes enterprise-scale cloud, data, AI, cybersecurity, and managed services capabilities. Together, they combine technology innovation with the delivery, security and scale required for transport and critical infrastructure environments.

With the August 2026 release of SensorMine version 4, natural language AI has been embedded within the platform, enabling authorities to query operational data in plain English, generate insights on demand and create customised dashboards without specialist technical skills. This makes advanced analytics more accessible to operational teams, helping authorities extract greater value from the data being collected.

“The combination of the latest IoT, AI and operational data is changing how infrastructure is managed. Organisations can now identify patterns and risks in near real-time that were previously difficult to see, helping them improve safety, performance and planning outcomes.”

Chris CookCEO, Altra

SensorMine: Turning Operational Data into Actionable Intelligence

By transforming camera feeds and operational data into meaningful information, SensorMine enables organisations to move from reactive incident reporting to proactive, evidence-based decision-making.

The solution combines AI, IoT and video analytics to deliver near-miss detection, traffic analytics, and road-safety insights through a single platform. It includes proven data models and deploy-ready dashboards providing an end-to-end approach to data capture and analytics without requiring additional development. This has been deployed for roads; at intersections, level crossings, and school crossings to measure active road users including vehicles, cyclists, and pedestrians.

Designed for rapid deployment and cloud agnostic, SensorMine meets stringent security and privacy requirements and can ingest feeds from existing cameras. Alternatively, self-contained portable cameras can be provided and deployed in about an hour.

Applications of AI-Driven Road Safety Technology

SensorMine has been successfully deployed by a number of local governments in partnership with the RSC and LAB3 to improve road safety outcomes.

The first initiative was for the Shire of Northam and the Near Miss pilot, where the project combined portable cameras with SensorMine's real-time analytics to identify and assess near-miss incidents involving vehicles and pedestrians. The solution provided insights into traffic behaviour, congestion patterns and areas of elevated safety risk, helping build a more detailed picture of how roads were being used. The pilot demonstrated how AI-driven road safety technology could generate actionable intelligence for transport planning, enabling local governments to make more informed decisions about infrastructure investment and community safety.

Building on the success of the Northam pilot, similar SensorMine deployments have since been undertaken with other local governments, including the City of Vincent and the City of Stirling. These initiatives have continued to demonstrate how AI can help agencies better understand road network performance, identify emerging safety risks and support evidence-based decision-making.

SensorMine Supporting the Fremantle Traffic Bridge Project

Continually advancing in capability, SensorMine is now also being used to support major infrastructure projects where road closures and traffic detours can place additional pressure on surrounding transport networks and create new safety challenges.

In 2026 Main Roads WA deployed SensorMine as part of the construction of the new Fremantle Traffic Bridge. With the closure of the existing bridge expected to significantly affect traffic flow, the project required reliable, real-time intelligence to support operational decision-making. SensorMine helps address gaps in traffic visibility by providing deeper insights into driver behaviour, freight movements and network performance, enabling stakeholders to respond more effectively to changing conditions.

Over a three-month period, SensorMine provided 24/7 insights into near misses, freight movements, traffic volumes, vehicle classifications and network usage patterns across the Fremantle Port precinct.

Cameras were deployed at strategic locations including Tydeman Road, Port Beach Road and Stirling Bridge, providing Main Roads WA with a continuous view of network performance. The extended monitoring period enabled teams to understand how changing conditions and freight movements influenced road-user behaviour, helping keep the project on schedule.

What Is the Future of AI in Transport and Infrastructure?

By helping organisations identify risks earlier, understand behaviour patterns and respond with confidence, platforms such as SensorMine are enabling a more proactive approach to transport and infrastructure management.

For LAB3 and Altra, the goal is simple: help organisations make better use of data, improve safety outcomes and build more resilient transport networks for the communities they serve.

“Near misses are often warning signs that something needs to change. AI gives transport authorities the ability to identify risks earlier, understand behaviour across the network and make decisions that improve safety before a serious incident occurs.”

Kerry PurcellCEO, LAB3