Scientists Outwit Poachers With The Help Of Artificial Intelligence
In the last hundred years, tiger populations across the world have decreased; from a 60,000 population, only 3,200 were left in the wild by 2010. Poaching has been the main reason for this unnerving drop in numbers, as these amazing big cats are killed for their skins, poaching, medicine, or even trophy hunting, driving them to near extinction. As if it's not enough, poachers have also been attacking other large animals such as elephants and rhinos, all of which play crucial roles in the ecosystems where they live.
— Carles Dijous (AAlb) (@carlesdijous) April 23, 2016
To keep these animals safe, Human Patrols roam around parks to keep poachers off protected areas, but according to Science Daily, agencies still have a limited resource for them. Fortunately, the National Science Foundation and the Army Research Office - with the help of researchers and conservationists from the US, Singapore, the Netherlands, and Malaysia - have found a way to use artificial intelligence and game theory to solve poaching, illegal logging, and other environmental problems worldwide.
An NSF-funded team from the University of Southern California has been building on the idea of "green security games" or the application of the game theory to wildlife protection. The game theory, with its use of mathematical and computer models to analyze rational decision-makers to help predict behavior and plan approach, can be used to aid in wildlife protection. With the help of AI, the Protection Assistant for Wildlife Security (PAWS) can used data of past patrols to learn patrol planning, eventually improving its system to deal to more observations of poaching activities, among other things.
Fei Fang, a PhD candidate in Computer Science from the University of Southern California and a member of the team, shared the importance of protecting natural resources and wildlife populations, saying that their work on PAWS is only just the beginning of improving ways to combat illegal poaching.