Artificial intelligence (AI) may seem a long way removed from the natural world. But a new collaboration between WWF and tech giant Intel, announced yesterday on Global Tiger Day, is harnessing the power of AI to help protect wild tigers and their habitats.
Back in 2010, WWF together with 13 tiger-range governments launched the TX2 goal — a commitment to double the number of tigers in the wild by 2022, the next Chinese year of the tiger. While there are many aspects to achieving this goal, one crucial component is being able to monitor trends in tiger distribution and population numbers. Essentially, the more we know about how many tigers there are and where, the better we’re able to protect them.
Camera traps, triggered to take a photo or to record a video stream when they detect movement, are a crucial tool for monitoring tigers. But today’s advanced camera traps routinely collect tens of thousands or even hundreds of thousands of photos, and as their use has expanded, analysing the masses of data they collect has become an almost impossible task. And that’s where AI can help.
While camera traps can easily be triggered accidentally — by a moving branch, for example — Intel’s technology will identify whether a tiger and or another wild animal is really in the shot before the image is captured. This means that rather than having to sift through numerous blank frames, every image is useful. Using cloud technology, the image can then be sent instantly to a monitoring centre, so we can see what’s happening in near-real time.
But where the technology really comes into its own is in helping identify individual tigers. Like human fingerprints, each tiger has a unique pattern of stripes. When we capture an image of a tiger, we need to compare these patterns against a catalogue of other images in an attempt to identify the individual. This can take many hours of work, which adds up to a significant cost.
A team of Intel scientists is working to “train” the technology to recognize individuals from their stripe patterns. It will also be used to identify images of different prey species — monitoring these is another important aspect of tiger conservation.
The information is attached, or “tagged”, to each image, making the job of analysing the data much quicker and more accurate.
We’re looking forward to working with Intel to explore more ways that technology can help advance wildlife monitoring and nature conservation. WWF is already using innovative technologies to take conservation to new levels, from infrared cameras and unmanned aerial vehicles to blockchain and satellite image processing.
The AI technology will be piloted in Huangnihe National Nature Reserve in northeastern China. This is part of a critically important landscape for the recovery of the Amur tiger. Today, around 550 Amur tigers remain in the wild, with the vast majority living in the Russian Far East.
As recently as 2010, there were thought to be no more than 20 wild Amur tigers in China. All of them had crossed over from Russia, and none were breeding females.
But in 2014, our camera trap footage revealed a mother tiger and two cubs in the Jilin Wangqing Nature Reserve — evidence that Amur tigers are breeding in China again. There’s great potential to increase their numbers, and President Xi Jinping has shown a personal interest.
To enable this, China is greatly increasing the extent of its protected areas — including creating a new Amur Tiger and Leopard National Park covering 14,600km2, which is 60 per cent larger than Yellowstone National Park in the US.
Better protection of northeast China’s forest habitats doesn’t just benefit tigers. It also helps conserve countless other species, as well as storing carbon, protecting vital watersheds and maintaining the way of life of communities who live in and around the forest.
Doubling the number of tigers in the wild is an ambitious goal, and we need all the support possible to make it happen. Applying AI to practical solutions is good news for tigers, for people and for nature.
— — — — — — — — — — — — — — — — — — —
By Alistair Monument, Forest Practice Leader, WWF, and Maggie Kinnaird, Wildlife Practice Leader, WWF