Developing a Braitenberg Controller for Autonomous Navigation in Duckietown

Session Number

1

Advisor(s)

Dr. Matthew Walter, Toyota Technological Institute at Chicago, Robotics Intelligence through Perception Lab

Location

A119

Discipline

Engineering

Start Date

15-4-2026 10:15 AM

End Date

15-4-2026 11:00 AM

Abstract

The Duckietown Project was created by the MIT graduate class in 2016, and has transformed into a worldwide program. The Duckietown platform uses a robot consisting of cameras, actuators, and April Tags to navigate a modular cityscape. This research focuses on developing autonomous robots that integrate internal and external sensors with image processing to interpret their surroundings. The research aims to develop an algorithm enabling the Duckiebot to detect obstacles within the Duckietown environment. A Braitenberg vehicle-inspired controller is implemented to connect sensor inputs directly to motor outputs. This approach allows the Duckiebot to react dynamically to the environment by adjusting wheel speeds. Through this research, we utilize Docker, Linux, Python, and machine learning to detect colors and objects.

Share

COinS
 
Apr 15th, 10:15 AM Apr 15th, 11:00 AM

Developing a Braitenberg Controller for Autonomous Navigation in Duckietown

A119

The Duckietown Project was created by the MIT graduate class in 2016, and has transformed into a worldwide program. The Duckietown platform uses a robot consisting of cameras, actuators, and April Tags to navigate a modular cityscape. This research focuses on developing autonomous robots that integrate internal and external sensors with image processing to interpret their surroundings. The research aims to develop an algorithm enabling the Duckiebot to detect obstacles within the Duckietown environment. A Braitenberg vehicle-inspired controller is implemented to connect sensor inputs directly to motor outputs. This approach allows the Duckiebot to react dynamically to the environment by adjusting wheel speeds. Through this research, we utilize Docker, Linux, Python, and machine learning to detect colors and objects.