Motor Learning in a Robot: Testing Competing Ideas from Neuroscience

Overview

Objectives/Goals Movements need to be accurate for humans and animals alike. Therefore, the method used by the nervous system to generate these movements is important. There are many aspects of motor control, such as direction, speed, and distance of movement. This experiment examines how to control the distance of a movement. Neuroscientists have proposed two possible methods for accurately controlling the motor commands used to generate the distance of a movement: timing and corollary discharge. The timing method uses timers in the brain to measure the amount of time from the start of a movement to determine when to stop. Alternately, the corollary discharge method uses a copy of its own motor command to estimate the distance traveled in a movement in order to decide when to stop. Methods/Materials In order to perform a realistic test of these two methods, I needed to 1) implement each method in a separate program to control a robot, and 2) test the programs in a challenging task. Each program was written in RCX code, and downloaded into the LMRIS 2.0 robot, called Roverbot. The challenging task was to make it learn from a single experience to stop before it hit an object. During the learning experience, the Roverbot traveled forward on a track until hitting a wall. During future trials, it traveled a distance that was based on the motor control method. Results After testing the two methods in a robot, I found that corollary discharge is a more consistent method than timing when there are random variations in the motor#s commanded speed. Conclusions/Discussion Although neuroscientists propose that both methods are used by the brain to control the distance of movements, this data supports the corollary discharge method by showing that movements would be more accurate if the brain used corollary discharge rather than timing. The data also suggests that corollary discharge would be helpful in building robotic devices and possibly prosthetic limbs.

Competition history

• CSEF 2007 Mathematics & Software · Entry J1311

Resources

• CSEF project summary ↗

Related projects

Closest projects by meaning, across every fair and year in the corpus.

Browse more like this

• All Computational Sciences projects
• CSEF Computational Sciences projects
• Every CSEF 2007 project