Artificial Intelligence for Wildlife Conservation
Using AI and algorithms in the fight against poachers.
University of Southern California, United States
With climate change, deforestation and poaching, conversation efforts are vital to protect our planet’s flora and fauna. The illegal wildlife trade is now the fourth most lucrative transnational criminal activity after drugs, arms and human trafficking, according to several wildlife charities including the World Wildlife Fund (WWF) and the International Union for the Conservation of Nature. It is said to be worth up to $23 billion per year.
In recent years a new tool has emerged in the fight against poachers, and that’s artificial intelligence – the ability of a machine or software program to ‘think’ and learn. The Teamcore Lab at the University of Southern California's Center for Artificial Intelligence in Society is working on two projects that use artificial intelligence and other technologies to outsmart poachers.
One approach is to augment the use of longwave thermal infrared cameras mounted on drones to detect poachers at night and report them to rangers. Monitoring live feeds is an arduous task and so Professor Milind Tambe and his colleagues developed SPOT (Systematic POacher deTector), an innovative application that automatically detects poachers and animals in near real time. "It essentially raises an alarm for a human being to then say OK I see a poacher in this video let me now call a ranger in order to try and intercept the poacher," said Prof Minde in an interview with the Robohub podcast.
The other AI-driven method developed by the Teamcore Lab is called PAWS, short for Protection Assistant for Wildlife Security. A machine learning algorithm analyses past poaching data to predict where poachers might next set their traps allowing rangers to detect traps and remove them before they can harm wildlife.
With a game theory framework that uses a branch of mathematics focused on the interaction between two or more participants, the system generates GPS locations for rangers to patrol. PAWS anticipates the actions and reactions of poachers and rangers and recommends patrol routes based on the probability of attacks and predicted poacher behavior.
With PAWS, forest areas are divided up into grid squares with each square being a target where a poacher may strike. Then the system calculates how frequently to patrol each grid square. When poachers attack a target, data is collected which is used to predict where they might strike next, allowing rangers to improve their patrols.
In early field tests, the system notched up a number of successes. For example, rangers tested PAWS at Srepok wildlife sanctuaries in Cambodia. Twenty-four rangers patrolling areas identified by PAWS discovered more than 1,000 snares from mid-December 2018 to late January 2019 more than double before the introduction of the AI software.
"PAWS helped us move from a reactive position to a more proactive approach," said James Lourens, law enforcement technical adviser for the Srepok Wildlife Sanctuary "I've been really happy with the results."
A New Era in Conservation
AI technology is ushering in a new era in conservation and is being used not only to combat wildlife crime but also to monitor animal populations, harnessing valuable data to know more about them.
Next Story »