Matter launches AI stormwater sensors for flood detection

Matter has introduced an AI-integrated sensor system for stormwater network management, with the technology already in use in the City of Monash in Victoria.

The system combines sensors that measure light and water with agentic artificial intelligence. According to Matter, it can identify blockages, debris and damage before water begins to flow through drains and pipes. This marks a shift from conventional monitoring systems, which typically alert operators only after changing water levels suggest a problem may already be developing.

Flooding events in Australia have increased scrutiny of how councils and infrastructure operators maintain drainage networks. Much of that work remains labour-intensive and costly, with crews often responding only after heavy rain exposes blocked pits, sediment build-up or damaged assets.

Matter director Len Luxford said the company sees scope to shift maintenance planning away from emergency response and towards earlier intervention based on sensor data and pattern recognition.

"At the moment, stormwater management is largely reactive. Current sensor systems send back data as water starts to flow, and it's not until water levels rise or water isn't draining as quickly as it should that a blockage or other issue is identified," Luxford said.

He said measuring light gives operators another way to detect changes inside a pipe or pit before rainfall creates visible flow changes.

"By integrating agentic AI into sensors that measure light as well as water, blockages, damage and other issues can be identified well before the first water starts flowing," he said.

Matter describes the setup as a recursive system that can identify trends and make evidence-based predictions about likely issues in parts of a network. In practice, that could help asset owners decide where to send maintenance crews first, rather than inspecting sites in a fixed sequence or waiting for field reports.

"Suddenly stormwater management can be proactive, creating operational efficiencies, reducing costs and potentially minimising the risk of flooding events," Luxford said.

Current deployments

The technology has already been deployed in the City of Monash and is being prepared for installation in Yarra Trams rail pits. These settings suggest Matter is targeting both municipal drainage systems and transport infrastructure, where water ingress and blocked drainage can disrupt operations.

Rail pits are a notable use case because drainage faults in transport corridors can have knock-on effects beyond local flooding, including maintenance access issues and interruptions to network operations. A system that flags likely sediment build-up or localised failure before a major weather event could allow works to be scheduled earlier and in lower-risk periods.

Luxford said the approach could also improve worker safety in parts of the network that are difficult or hazardous to access, including pits near busy roads and highways.

"We no longer have to wait for water to start flowing before an issue is identified. This system can predict, with evidence-based accuracy, where the blockages are ahead of time. This allows works to be prioritised and resources, including staff, to be deployed more efficiently," he said.

Broader use

Matter argues that predictive monitoring could support more localised maintenance planning by drawing on historical information and site-specific conditions. Weather patterns, topography, nearby waterways and the level of urban development can all affect how stormwater networks behave, making broad seasonal assumptions less useful than location-specific data.

The company says an agentic system can also compare readings across clusters of sensors, which may help if one sensor fails and nearby devices still provide enough information to assess conditions. It adds that network data could be used alongside other inputs, such as weather applications, to give operators a broader view of conditions across an area.

Cost is another part of the pitch. Luxford said older stormwater sensors have often been too expensive for widespread deployment, limiting visibility across larger networks and in more remote locations.

"We can stop reacting and start predicting, and that will be an absolute game-changer for the industry," he said.

He said smaller, cheaper devices could allow councils and businesses to place more sensors across their networks, extending monitoring to areas that have previously been harder to justify on budget grounds.

"Sensors of the past have been expensive, meaning councils and businesses have been limited by budget in how many they could have across stormwater networks," Luxford said.

"By creating technology that is smaller and more affordable, extra sensors can be deployed, providing an enhanced view of the network and improving operational efficiency and performance.

"Combined with conservative battery life of around a decade, we suddenly have the ability to monitor areas of stormwater networks we haven't been able to in the past, such as remote and regional areas. The opportunity this will provide moving forward is significant," he said.