I work with wildlife biologists and government researchers around the world to improve current wildlife survey methods. I focus my work on marine mammals and seabirds, like sea lions, albatross and penguins, but have surveyed a wide range of species including guanaco and flamingos. Additionally, I have helped develop and build wildlife conservation drone programs around the world by providing my knowledge of drones and post-processing software. More recently, I have been involved in several projects counting wading birds, as well as large colonies of elephant seals, and will be using 'automated' post-processing analysis like BirdDetector and ImageJ to count the individuals in the large orthomosaics.

 

Most large colonies are monitored by sampling a part of the colony and extrapolating to estimate the entire colony. This is due to large colony sizes, difficult and hard-to-access terrain, etc. Using drones, surveys of the entire population can now be conducted to get a more accurate estimate for the colony size. Although researchers must still process the images and count the individual animals on a computer, recent advances in machine learning, deep learning, and computer vision have allowed automated counting models to be built with a high level of accuracy. This makes before/after surveys more accurate; whereas different pairs of researchers counting the same colony may inhibit human bias, the mission planning software and counting model will be the same each year, creating a more reliable and accurate understanding of population dynamics. 

 

I recently graduated from the Bren School of Environmental Science & Management, with a Master of Environmental Data Science, where I gained skills in programming and machine learning. I hope to improve these skills and gain a better understanding of object detection and neural networks. My goal is to integrate this with my drone skills so I can analyze the drone images I take so that I can provide actual analytics to the clients I work with, not just large orthomosaics. 

 

I primarily work with quadcopters, such as the DJI Mavic 2 and the Autel Evo II, but also work with fixed-wing drones like the SenseFly eBee X.  I have just begun working with the Quantum Trinity Pro, a VTOL fixed-wing drone, which has proved to be very effective covering large areas in the field and along coastlines. Thermal sensors have proved useful when trying to see birds through shrubs, and multispectral sensors have been useful for vegetation health surveys.