Friend and colleague, Richard Roberts, has offered to give us an glimpse of his recent work on learning autonomous robot behaviors/controllers from human operators:

While working on the DARPA LAGR project, we found it exceedingly difficult to tune our reactive behaviors to work well in cluttered and patchy environments. Either obstacle avoidance was too sensitive and the robot would not drive through gaps, or it was too aggressive and the robot would collide with obstacles. Of course, we could have made the behaviors more and more complicated, introducing more parameters to tune, but we wanted an easy way to have the robot “just do what I say!”  Thus, we developed a system for interactive, on-line training of behaviors with a remote control. The user flips a switch to training mode, and drives the robot how they would like it to drive, then flips the switch back to autonomous to test the behavior.