
Being a ranger in Cambodia is an incredibly risky job, with four rangers being killed by poachers just last year. It is difficult to know when and where to patrol in order to catch poachers in the act- until now.
Researchers at the University of Southern California Viterbi School of Engineering have developed a novel way for rangers to decide on a safe and effective patrol route.
The “Protection Assistant for Wildlife Security” (or “PAWS”) is a predictive AI software, employing mathematical models and game theory.
It is used in combination with the digital tracking tool already employed in natural reserves across the world, the “Spatial Monitoring and Reporting Tool” (or “SMART”). This software enables rangers to collect data from their patrols, which can then be analysed – but SMART is unable to predict the poaching hot spots, which is the advantage of using PAWS as well.
And this combination has been proven to be effective. Since last year, PAWS has been trialled in Srepok wildlife sanctuaries. Twenty-four rangers patrolled areas identified by PAWS finding more than 1,000 snares between mid-December and late January – more than doubling their findings before the software was employed.
James Lourens, law enforcement technical adviser for the Srepok Wildlife Sanctuary said that he has been very happy with the results: “PAWS helped us move from a reactive position to a more proactive approach”.
Field testing is expected to continue into the summer, with data from December and January being incorporated into the model to give a richer understanding and maximise outcomes.
https://www.bbc.co.uk/programmes/p07kkc53
https://magazine.viterbi.usc.edu/spring-2019/features/srepok-wildlife-sanctuary/









