Given that our project is a multi-year, multi-student work of engineering, we know our progress can be hard to follow. We've created this page to show you our progress in an organized way.
The Cornell Autonomous Bicycle Team develops a robotically-stabilized bicycle with the goal of balancing "better" than any other autonomous bicycle. Others have developed autonomous bicycles, often balancing using gyroscopes or reaction wheels. Gyroscopes and reaction wheels are heavy and energy inefficient. Our bicycle balances only by steering the front wheel. This is similar to how humans balance bicycles. We are working to make our bicycle navigate autonomously around Cornell's campus. We aim to make our bicycle robust to disturbances (wind, bumps, etc).
Our bicycle can autonomously balance! This semester (Spring 2018), we developed and tested new balance controllers, both linear and nonlinear, in simulation and on our prototype bicycle. We took the first steps to implement a computer vision system: we conducted preliminary tests of obstacle detection and localization algorithms using a stereo camera. We integrated more sensors into our position estimation algorithm. This improved the bicycle's ability to know its location. Next semester, we hope to implement our navigation algorithm on our physical bicycle to make our bicycle fully autonomous.