In National Parks around the world, rangers are fighting a constant battle with illegal poachers, patrolling hundreds, and sometimes even thousands of miles in an attempt to combat the devastating population decline of some of the world’s most endangered species.
It’s an incredibly challenging task, and a lack of resources means many poachers go uncaught. However, a new Al-assisted approach could lead to a significant increase in captures, by optimising patrol routes.
“In most parks, ranger patrols are poorly planned, reactive rather than pro-active, and habitual,” according to Fei Fang, a PhD candidate at the University of Southern California (USC) computer science department.
Known as PAWS, or Protection Assistant for Wildlife Security, the Al system is responsible for planning daily-changing patrols to follow routes that offer the highest chance of capturing poachers.
This is achieved by incorporating complex terrain information – including topography – alongside data on past patrols and evidence of poaching. PAWS even considers the natural migratory paths of animals, which are popular poaching targets.
Although the system is still in development, it has already led to more identification of poacher activities per kilometre.
“This research is a step in demonstrating that AI can have a really significant positive impact on society and allow us to assist humanity in solving some of the major challenges we face,” said reset leader Milland Tambe, professor of computer science and industrial and systems engineering and director of the Teamcore Research Group on Agents and Multiagent Systems.
Each day, PAWS calculates a new patrol route that minimises terrain changes – and therefore ranger fatigue – while covering as many likely poaching hotspots as possible.
“We need to provide actual patrol routes that can be practically followed,” said Fang. “These routes need to go back to a base camp and the patrols can’t be too long. We list all possible patrol routes and then determine which is most effective.”
The system even randomises rates to ensue poachers cannot spot patterns and thus avoid areas where they are more likely to be caught
“If the poachers observe that patrols go to some areas more often than others, then the poachers place their snares elsewhere,” said Fang.
PAWS was first developed in 2013, but has been improved significantly in the interim thanks in part to its ability to learn over time.
It was trialled in both Uganda and Malaysia in 2014, and has been formerly used in Malaysia since 2015. However, it has been improved further since with the incorporation of CAPTURE, or Comprehensive Anti-Poaching Tool with Temporal and Observation Uncertainty Reasoning, which has further improved the system’s ability to predict poacher behaviour.
“There is an urgent need to protect the natural resources and wildlife on our beautiful planet, and we computer scientists can help in various ways,” Fang said. “Our work on PAWS addresses one facet of the problem, improving the efficiency of patrols to combat poaching.”