How do animals process the tremendous amount of information coming
from their senses, in time to take appropriate actions with their
muscles? This type of robust sensorimotor processing is still
difficult to replicate in artificial systems even with the latest
computers, sensors, and actuators. I will describe some of my lab's
recent work on algorithms used to construct real-time robotic systems
for navigation, locomotion, and other autonomous behaviors. I will
also discuss our recent theoretical work in novel machine learning
techniques for dimensionality reduction, and how they can be applied
to such systems.