Robot Safe Control for Navigation
Robotics, Safe Control (Potential Field Approach, Safe Set Algorithm, Control Barrier Function), Navigation
Description
Motivation of this course:
Deploying mobile robots ubiquitously requires that they safely and reliably accomplish navigation tasks in crowded and dynamic real world settings. These settings are challenging since the robot system is expected to plan online, handle the uncertainty, and establish safe actions to avoid multiple moving agents.
Course Content and Overview:
On this course we will primarily use Safe Set Algorithm (SSA) as safe controller to keep monitoring and modifying actions for safety-critical tasks like autonomous driving and human robot interaction. You will learn the problem of naive safety strategy, how safety index solve the problem in naive method, and how to add constraints on control to guarantee safety. Besides the mathematics of safe control algorithms, I'll also show you how to code a safe controller and implement it in a challenging environment which simulate the real-world crowded and dynamic scenarios. The coding program would help you fully understand this algorithm and may apply it into your own applications.
Requirements:
This is a beginner robotics and control course - if you've never learned robot navigation before, this is a great starting place. No necessary control or robotics knowledge requirement, but require basic knowledge of math derivatives and some knowledge about python would be a great advantage.
What You Will Learn!
- understand the application of safe control
- know the limitation of naive safety strategy
- understand the math behind safe control
- know how to code the safe control for navigation in crowded dynamic environment
Who Should Attend!
- People who are interested in robotics control
- People who want to guarantee safety in robot navigation