LGAQ Lab

AI-powered bark logging

Barking dog complaints have long been a no-win issue for councils. Complainants endure the hassle of manual logs, dog owners feel unfairly targeted, and councils are left conducting unreliable spot checks. To solve this, Mackay Regional Council partnered with the LGAQ Lab to co-design an AI-powered solution that automates the entire barking detection process. The result is a low-cost, fully offline device that isolates dog barks from other neighbourhood noises using a custom-trained AI model. When the device is loaned to residents, the device automatically logs bark events with an 80% confidence threshold, creating accurate, continuous records without human intervention.

To further streamline the process, the LGAQ Lab developed an AI tool that reviews thousands of logged events within seconds. The system compares barking data against the council’s own noise legislation and generates a detailed compliance report. This report identifies potential breaches and highlights barking patterns, empowering councils with fast, evidence-based insights. The LGAQ Lab’s AI bark detection system reduces frustration for everyone involved, while giving councils a secure, data-driven, and efficient way to manage neighbourhood noise complaints.

Early warning flood alerts

The Bloomfield Bridge is a shared piece of infrastructure by Wujal Wujal Aboriginal Shire Council, Douglas Shire Council and Cook Shire Council. During ex-Tropical Cyclone Jasper, large portions of these communities were destroyed, including the traditional flood warning siren that alerted residents to rising floodwaters. Wujal Wujal Aboriginal Shire approached the LGAQ Lab to seek assistance in exploring the replacement

The LGAQ Lab designed a solution at 5% of the cost of the traditional flood siren, using LiDAR sensors to take river height readings every five minutes with 1mm of accuracy. When floodwater was detected 50cm below the bridge deck, an SMS is sent to subscribed residents. The LGAQ Lab took this further by developing a predictive model that uses machine learning to compare river heights with a century of tidal data to give councils an indication of river heights for the coming six days.

Predicting mosquito outbreaks

Mosquito outbreaks occur in many coastal areas, with many councils undertaking significant prevention methods only for the community to perceive the efforts as being insufficient. Mackay Regional Council sought the LGAQ Lab’s assistance in anticipating outbreaks before they occur, allowing for more effective prevention methods to be undertaken and more proactive communication with the community.

The LGAQ Lab designed a predictive model that uses AI to anticipate outbreak areas using decades of mosquito population counts, locational data, treatment records, temperatures and rainfall, confidently informing council of potential outbreaks up to four weeks before they occur. This allows council to undertake more effective prevention methods in at-risk areas and advise the community of personal precautions they can take to protect themselves.

AI-powered flooded road detection

When several camera hardware providers exited the market, Carpentaria Shire Council was left unable to access images from its existing road cameras to assess flood conditions. Council officers had to manually check an FTP server and, when images were unclear, travel more than five hours each way to inspect sites in person. This approach was time-consuming, inefficient, and at times unsafe. To address this, Carpentaria partnered with the LGAQ Lab to co-design an AI-powered flood detection system that could automate the entire process.

The LGAQ Lab developed a cloud-based, hardware-agnostic AI model that analyses road camera images captured every 15 minutes. When flooding is detected, or when a road clears, the system instantly emails council officers with the image, location, and detection confidence level. This solution removes the need for manual review, dramatically reduces travel time, and delivers near real-time flood assessments. Most importantly, it keeps council officers safe while improving the efficiency and reliability of flood monitoring across remote regions.