Learning and planning under uncertainty:

Wildlife poaching fuels the multi-billion dollar illegal wildlife trade and drives countless species to the brink of extinction. To help rangers prevent poaching in protected areas around the world, we developed the Protection Assistant for Wildlife Security (PAWS). We made technical advances in multi-armed bandits and robust sequential decision making through reinforcement learning, with research questions arising from challenges in the field. We also discussed how to bridge the gap between research and practice, presenting results from field deployment in Cambodia and large-scale deployment through integration with SMART, the leading protected area management software system used by more than 1,000 wildlife parks worldwide.