In a groundbreaking initiative to address the endangered status of koalas in northeastern Australia, researchers led by Professor Jun Zhou from Griffith University have unveiled a pioneering approach. Facing the challenge of accurately counting the elusive koalas in the wild, the team has integrated artificial intelligence (AI) with camera technology to establish a continuous monitoring capability. This innovative solution not only promises to provide a more accurate count but also offers insights into the impact of urbanization on koala populations.
Monitoring Koalas with AI
In July 2021, Zhou and his team embarked on a project deploying 24 cameras strategically placed at “koala crossing locations” in the Brisbane and Redland city council areas. Triggered by koala movement, these cameras automatically recorded footage, forming the basis for training an AI system.
The collaboration between Griffith researchers and local community koala monitoring groups aimed to not only spot koalas but also achieve individual recognition. This breakthrough enables continuous monitoring over larger areas, with the potential to place 100 cameras in a one square kilometer research region, operating 24/7.
Highlighting the importance of the AI technology, Zhou underscores its significance, stating that their AI possesses the capability to discern and scrutinize videos, distinguishing whether they feature koalas and identifying specific individuals. This, he explains, facilitates the comprehensive tracking of all koala movements. This capability is crucial not only for tracking existing koalas but also for identifying new individuals in the region.
Urbanization threatens Koala numbers
While the AI-assisted monitoring brings hope for better conservation efforts, early estimates from the Redlands project reveal a grim reality. Dr. Douglas Kerlin, co-researcher, points to urbanization as the primary driver of the decline in koala numbers in Southeast Queensland. The encroachment of urban areas into their natural habitat poses a significant threat to koala populations, underscoring the urgency of effective conservation measures.
The team, buoyed by the success of the initial trial, has secured a grant from the Queensland Government to expand the project to 10 local government areas. Zhou envisions not only improving the accuracy of the technology but also fostering wider adoption. In recognizing the complexities involved, Professor Jun Zhou candidly admits that ushering the technology into a broader sphere will undeniably demand a considerable amount of time.
AI Koala and vital community participation
Recognizing the need for sustained efforts, Zhou emphasizes the role of local councils and koala conservation communities in deploying and maintaining the camera network. While the technology automates data capture, transfer, analysis, and reporting, human intervention remains vital.
Advocating for community engagement, Zhou underscores the value of involving local residents in the initiative, suggesting that, rather than the research team managing camera maintenance, the community could play a pivotal role in deploying and sustaining the camera network. This collaborative approach aims to ensure the long-term success of the AI-driven monitoring system and its integration into wider conservation practices.
The AI koala project represents a ray of hope for the endangered species, offering a technological lifeline in the face of urbanization threats. As researchers work towards refining the technology and expanding its reach, the involvement of local communities emerges as a crucial element in the journey to save the iconic koalas of northeastern Australia. As the AI Koala initiative continues to unfold, the intersection of technology and community engagement opens a promising chapter in the ongoing endeavor to safeguard these iconic marsupials from the challenges they face in their natural habitat.
