How AI is Being Used to Combat Poaching

Artificial Intelligence is helping tackle poaching problems, offering new hope for endangered species

As the human population booms (7.7 billion as of May 2019), we’re seeing a mass extinction of other species taking place. There are many contributing factors; disease, urbanisation, and climate change to name a few. All of these are synonymous with humans and the devastation we’re causing to the planet.

Poaching is playing a huge part in this. According to a recent article from WWF, the illegal ivory trade is at the highest it’s been for 20 years. A staggering 20,000 African elephants are killed each year, that’s 55 a day. This means that more elephants are being poached than are being born. Whilst steps are being made to combat the ivory trade by governments, it’s not just limited to poaching elephants for ivory. Compared to 60,000 a century ago, only 5,574 tigers now remain in the wild. On average, 3 Rhinos lose their horns to poaching a day. Shockingly, only 70 Amur Leopards remain in the wild due to poaching for their skin and bones. They also lose their habitat due to illegal logging.

Not only is poaching destroying species, but it has devastating effects on ecosystems and economies the world over. Biodiversity is so important in the conservation of our planet, and wiping out one species has detrimental effects.

I struggle with the reasons why poaching still exists in this day and age. Of course, the main driver being profit. Driven by wealthy individuals that believe ivory trinkets and skin rugs are more suited to being in their houses than on the animals themselves. Similarly, the medicinal properties of animals result in poaching. Since the 1500’s Rhino horns have treated various illnesses in Traditional Chinese Medicine. Spoiler alert: it doesn’t actually work.

For years, many charities and conservationists have worked tirelessly to combat these issues. Thanks to Artificial Intelligence, conservationists are now able to tackle these problems in a different way. Using this technology to get ahead of the poachers, we can now look to taking more preventative measures to poaching.

 

PAWS

PAWS (Protection Assistant for Wildlife Security) is a fantastic system doing just this. It uses Artificial Intelligence and Machine Learning to assess data and predict poaching activity. Reservations and land where protected species live are often too large to patrol all of the time. This means, unfortunately, poachers manage to kill these innocent creatures where they should be most protected.

By inputting data collected previously on poaching activity, data on the local areas and patrols, it uses an algorithm to generate new patrol routes that will be most effective in combatting poaching activity and hopefully reduce the number of fatalities to wildlife.

These patrol routes and outcomes are then fed back into the system, to keep the data new and also update new routes to take.

The main aspect in PAWS’ model is Computational Game Theory. They designed a ranger v poacher game to help further their understanding of strategies and logic poachers may use through an AI simulation. Similar to the techniques used in games like online poker.

With this model, they can generate patrol routes that lead to where poaching may be more likely to take place next, in order to tackle the problem head-on and protect these endangered species.

According to Reuters, in Uganda’s Queen Elizabeth National Park, rangers used this system and discovered 10 antelope traps, elephant snares, and a ‘betterscore card than without such technical assistance.’ However, there are obstacles that PAWS must overcome, such as putting rangers into too close quarters with violent poachers, and also problems with internet and mobile signal that can prevent the real-time use of the software out in the field.

Regardless of these issues, PAWS is a fantastic use of AI technology. Its contributions to preserving the lives of endangered species will hopefully see wider scale adoption soon.

Picture of Guest Writer

Guest Writer

Share my work: