Machine Learning in Biology - Teaching a Glider to Soar Like a Bird

Explore the intersection of machine learning and biology in this 25-minute talk that delves into teaching autonomous gliders to soar like birds. Discover how Gautam Reddy from the University of California, San Diego applies reinforcement learning techniques to enhance glider performance. Gain insights into Marr's levels of analysis, thermal soaring, and bird navigation in thermals. Examine the aerodynamics of gliding birds and the convective environment in the atmosphere. Learn about the reinforcement learning framework for creating optimally behaving agents. Understand methods for measuring vertical wind acceleration and velocity gradients. Follow the process of training gliders in the field and consider whether birds can sense vertical wind gradients. Uncover the fascinating applications of machine learning in improving autonomous flight capabilities.

Intro
Marr's levels of analysis
Thermal soaring
How do birds find and navigate thermals?
Aerodynamics of a gliding bird
Convective environment in the atmosphere
An optimally behaving agent: the reinforcement learning framework
Measuring the vertical wind acceleration
Measuring the vertical wind velocity gradients
Training the glider in the field
Can a bird sense vertical wind gradients ?