Project title: Developing hybrid robots for navigation
Dr. Cen Li

Hybrid robots are robots programmed using a two-layer approach, with the top layer performing the more sophisticated tasks, e.g., planning, behavior generation and sequencing, and the lower layer executing the more primitive reactive behaviors, e.g., obstacle avoidance.

This project involves programming the Khepera robot to navigate in an area modeled after a typical office cubicle environment. The Khepera robot is equipped with 6 build-in IR sensors and light sensors, as well as differential wheels with encoders. A typical task of the robot is to navigate from one office room to another office room specified by the user.

The robot should be able to load/process map of the environment, plan on an efficient path to go from the starting place to the goal place, form and sequence the behaviors to be executed, execute the behaviors (e.g., navigate hall, move through door), and update the map as necessary. The robot should also be able to perform re-planning when discrepancies are found between the map and the real environment.

Programming requirement: C/C++ and willingness to study the specifications of the hardware, learn the library functions provided with the Khepera robots.