Autonomous lane change of an F1/10 unit

This research project was a part of the SURF research program at California Institute of Technology. In this work we focus on implementation of localization and control systems of an F1/10 unit in order to perform autonomous driving in multi-lane traffic. We use lidar data in order to map and localize in the environment using SLAM. For this purpose we used the ROS package Hector Slam. The map generated online is illustrated in the video below.

For path planning and obstacle avoidance we use motion primitives to generate a trajectory. The generated trajectory is tracked using a PD-controller. The video below shows the F1/10 unit avoiding an obstacle in order to advance.

The work is presented in the paper Design and implementation of a controller to advance in multi-lane traffic similar to human behavior using control improvisation. The research was conducted at California Institute of Technology.