Australian scientists are on the verge of a revolutionary advance in detecting bushfires, utilizing cube satellites equipped with artificial intelligence to identify fires from space 500 times faster than traditional ground-based methods. This leap forward addresses the critical challenge of processing and compressing massive amounts of hyperspectral imagery onboard the more economical and compact cube satellites before relaying it to Earth, thus conserving vital time and energy.
This innovative approach leverages artificial intelligence to ensure that bushfires are identified earlier from space, even before they grow and generate significant heat. This early detection enables ground crews to respond swiftly, potentially saving lives and property. The SmartSat CRC, in conjunction with the University of South Australia (UniSA), has spearheaded this project, which uses cutting-edge AI technology to create an energy-efficient early fire smoke detection system for South Australia’s first cube satellite, Kanyini.
The Kanyini mission, a collaborative effort involving the SA Government, SmartSat CRC, and industry partners, aims to deploy a 6U CubeSat into low Earth orbit. This satellite is tasked with detecting bushfires and monitoring inland and coastal water quality. Equipped with a hyperspectral imager, the satellite sensor captures reflected light from Earth across various wavelengths, generating detailed surface maps for multiple applications, including bushfire monitoring, water quality assessment, and land management.
Traditionally, Earth observation satellites lacked the onboard processing capabilities to analyze complex images of Earth in real time. However, the team led by UniSA geospatial scientist Dr. Stefan Peters has developed a lightweight AI model capable of detecting smoke within the cube satellite’s limited processing, power, and data storage constraints. This model has dramatically reduced the volume of data downlinked to 16% of its original size while consuming 69% less energy and detecting fire smoke 500 times faster than conventional ground-based processing.
Dr. Peters emphasizes the importance of early smoke detection, stating, “Smoke is usually the first thing you can see from space before the fire gets hot and big enough for sensors to identify it, so early detection is crucial.” The team demonstrated the AI model using simulated satellite imagery of recent Australian bushfires, employing machine learning to train the model to recognize smoke in an image.
Most sensor systems collect vast amounts of data, only a fraction of which is critical to the mission’s purpose. Typically, this data is downlinked to the ground for analysis, consuming considerable space and energy. By training the AI model to distinguish smoke from clouds, the researchers have made the process much faster and more efficient. Using a past fire event in the Coorong as a case study, the simulated Kanyini AI onboard approach detected smoke and transmitted the data to the South Pole ground station in under 14 minutes.
Dr. Peters highlights the significant advantages of onboard AI over traditional ground-based processing, noting its potential not only for bushfire response but also as an early warning system for other natural disasters. The research team aims to demonstrate the onboard AI fire detection system in orbit by 2025 when the Kanyini mission becomes operational.
The ultimate goal is to commercialize the technology and deploy it on a CubeSat constellation, thereby contributing to early fire detection within an hour. The researchers have published their experiment’s details in the latest issue of the IEEE Journal of Selected Topics in Applied Earth and Remote Sensing.
This breakthrough in using AI for early fire detection represents a significant step forward in mitigating the devastating impact of bushfires. By harnessing the power of space technology and artificial intelligence, Australia could be better prepared to respond to future fire threats, potentially preventing the kind of widespread destruction witnessed in past fire seasons.
