Computational Sustainability for Everyone: Poaching the Poachers with PAWS

This post is by Zimei Bian. See her bio and contact information below.

Poaching is the illegal hunting, killing, or capture of wildlife. Preyed on by both trophy hunters and those that seek to profit from their extraordinarily-priced body parts, animals playing critical roles in our ecosystem have been pushed to near-extinction in recent years as their populations have dropped to unsustainably low numbers. To get an idea of the extent of the problem, the Black rhino population has declined by 97.6% since 1960, and up to 35,000 African elephants were killed just last year. To learn more about the motivations and impact behind illegal poaching, watch “Last Days”, a short film by Oscar-winning director Kathyrn Bigelow (of Last Days of Ivory), below:

(“Last Days”, a short film by Oscar-winning director Kathryn Bigelow)

To combat this global poaching crisis, a team of researchers at the University of Southern California led by CompSustNet Associate Director Milind Tambe are adding artificial intelligence (AI) to the mix. While human patrols at wildlife protection agencies serve as the primary form of protection for endangered species, the large sizes of reserves and limited resources give illegal poachers a significant advantage in avoiding capture. Funded by the Army Research Office and the National Science Foundation (NSF), Dr. Tambe’s team is working on the Protection Assistant for Wildlife Security, or PAWS, to assist conservation agencies in optimizing patrol routes that cover areas where poachers are most likely to attack.

(PAWS visual model – credit to

Developed in 2013, PAWS builds a behavior model that aims to predict poaching activities by analyzing existing patrolling and poaching data while accounting for natural routes with the most animal traffic. The software then generates randomized patrol strategies in the form of route suggestions for rangers to cover the potential problem areas while avoiding predictable patterns, taking into consideration the protected area’s existing terrain to minimize time and energy consumption. As more data is collected, the information is fed back into the PAWS system, allowing the software to “learn” and improve its strategies. You can watch the official video for PAWS here:

(PAWS video – Winner of Best Application of Artificial Intelligence @ AAAI Video Competition 2016)

The PAWS software is currently being tested. A field test was conducted in Uganda’s Queen Elizabeth National Park in 2014, and the study was presented at the AAAI Conference on Artificial Intelligence this past February. In the meantime, check out this poaching infographic by the African Wildlife Foundation and learn more about this incredible application of computational sustainability here.

Stay tuned for the debrief on other exciting projects in the Computational Sustainability Network!

Zimei Bian is a Computer Science  undergraduate at Vanderbilt University with a special passion for interactive storytelling and using tech for social good. In her spare time, she enjoys internet cat pictures and story-driven video games. The opinions expressed herein are Zimei’s and not necessarily those of Cornell University. You can reach Zimei at