Using Vision-guided Underwater Vehicles to Study Life in Our Oceans

The lives of marine animals play vital roles in influencing policy decisions ranging from conservation to food security and global health. Yet, many species behaviors’ remain elusive to science as we have little to no direct observations of them. I propose the use of vision-guided autonomous underwater vehicles (AUVs) as a scalable and opportunistic approach to tracking and monitoring marine animals and their behaviors in their natural habitats.

In particular, I show how to apply generalizable machine learning/data-driven approaches to AUV perception and controls, as well as animal behavior. These combined enable AUVs to follow any visible marine animal(s), including those that are rare or elusive, across any habitat or environmental conditions, while minimizing invasiveness or observation bias of animal behavior caused by the vehicle itself. Finally, I will discuss how these approaches and future opportunities in robotics and machine learning enable us to rapidly scale our in-situ observational capabilities to meet the spatial, temporal, and multi-species challenges needed by marine ecologists.

Speaker: Levi Cai, Colorado University - Boulder