We’re bringing new tech to ancient forests to improve the way orangutans are monitored

Amanda Aiza Amran, PhD student from Universiti Malaysia Sabah shares her field notes from Bukit Piton Forest Reserve in Sabah, Malaysia

Successful conservation requires good data. To protect orangutans in Borneo, it’s vital we know how many there are, and how the population is changing. But their shy and largely solitary nature, coupled with their remote and inaccessible habitat, means tracking these great apes is no easy feat. So instead we count their nests.

Orangutans make nests out of branches – high in the treetops – every night and sometimes during the day. By counting these nests, we’re able to estimate their numbers. Finding nests usually involves walking in straight lines through the forest looking up into the canopy – a long and laborious process. But now drones equipped with digital cameras are an increasingly cost-effective way to survey vast areas of forest. They can take hundreds, even thousands, of photos a day. The problem is that every image needs to be painstakingly scrutinised by an expert to spot the nests.

Bukit Piton Forest Reserve - Forest Restoration & Orangutan Conservation Project

Orangutan nest spotted at the planted tree
Bukit Piton Forest Reserve © WWF-Malaysia / Mazidi Abd Ghani

What if we could make this process more efficient? That’s where my research comes in. I’m harnessing the power of artificial intelligence (AI) to develop an automated way of detecting orangutan nests in drone images, to help conservationists quickly and accurately monitor their populations.

To understand how nests are surveyed, I visited Bukit Piton Forest Reserve in Sabah, Malaysia. While a drone took images from canopy level, I trekked into the rainforest on foot to look for orangutans and their nests with the help of WWF-Malaysia’s knowledgeable orangutan team. Back in the lab, I used the aerial images to train AI technology to learn the unique features of orangutan nests, such as their shapes and colour. From this, I’m developing a deep-learning model that will not only automatically detect and classify the nests in the photos, but will also be able to analyse them. It will help us understand the characteristics of the nests, such as how they’re constructed and how they differ between individuals or populations. We can also examine how the nests are used over time, how long they last, and how they change with weather patterns. By identifying common patterns in nest construction, we can better understand orangutan behaviour and their relationship with their habitat.

One of the biggest obstacles to protecting rare species is the difficulty of monitoring their populations, so AI has huge potential to revolutionise conservation efforts. Thanks to your support, WWF is able to embrace technological innovations and combine them with traditional survey methods to improve the accuracy and efficiency of monitoring work, assess the effectiveness of conservation actions, and make science-based decisions.

Thank you for your support helping to protect vulnerable wildlife for generations to come.

An orangutan with its baby sit on a branch facing the camera
© / Anup Shah / WWF