How artificial intelligence is revolutionising conservation
Most of us can now access artificial intelligence (AI) on a daily basis by using Siri on our iPhone. What was once pondered about in books, and imagined on the screen, is now available at our fingertips. And yet it seems all we can think to do with it, is prank it with silly questions.
But while the rest of us are satisfied playfully interacting with the magic of AI, one Deakin researcher is exploring how we might use this technology to conserve protected environments and the species that inhabit them.
A better mouse trap
A three-man Deakin research team led by Professor Yong Xiang, Associate Head (Research) of the School of Information Technology, is developing a technical system that will record and circulate crucial data on animal tracking in nearly half of Victoria’s national parks. This data has never before been available in such detail, and has the potential to create revolutionary new park management processes.
Tracking animals is key to understanding the health and habits of a national park’s ecosystem. Traditionally, Parks Victoria, the body responsible for management of Victoria’s parkland, used a network of sensor cameras set up throughout major national parks to monitor animal movement.
But more frequent temperature changes, storms and rugged conditions have caused the cameras to be set off by random events, resulting in thousands of irrelevant images being taken. ‘Many of the photos contained no animals at all and all the photos had to be sorted manually, which was extremely time consuming,’ says Prof. Xiang.
Deakin’s system will utilise machine learning and pattern recognition to detect the capture of animals within the sensor camera images. As the technology learns, it will also be able to recognise what breed of animal has been caught.
Prof. Xiang explains that this will be a massive step forward to recording accurate and useful data. ‘It will make a huge difference for us when we can remove the laborious step of trawling through thousands of photos,’ he says.
A new normal in preservation practices
The Deakin team’s effort represents a growing trend in modern tech being utilised to protect ecosystems around the world. Protection Assistant for Wildlife Security (PAWS) is a newly developed AI that collects poaching activities and behaviour data, to better inform animal protection patrols. The technology makes informed predictions about poachers’ movements and allows authorities to stay one step ahead of dangerous poachers.
Meanwhile NASA, which previously used satellite data to map bushfires on a macro level, is developing a new AI interface to better inform firefighters in the middle of fighting dangerous blazes. AUDREY, the Assistant for Understand Data through Reasoning, will be able to send real-time updates about the temperature, speed and threat of a fire to head mounted displays of firefighters. This process ensures the split-second decisions that determine life or death for first responders are made based on huge amounts of data humans couldn’t possibly analyse. In this way, artificial intelligence ensures not only the safety, but strategic soundness of firefighters potentially in harm’s way.
Professor Yong Xiang
Associate Head of School, Faculty of Science, Engineering and Built Environment, Deakin University
- Don’t miss