Control Barrier Function Implementation in C++
Problem Statement
Autonomy operating in safety critical environments requires guarantees on safe operation to be broadly adopted in society. Control Barrier Functions (CBFs) are one approach to provide these guarantees on broad classes of control systems. The CBF toolbox is a basic implementation of these techniques that provides a baseline for future autonomy development within the community.
Technology Solution
This package implements the CBF core concepts for standard control system dynamics to ensure safety of autonomous systems in C++ programming language
Value Proposition
Approach enables theoretical safety guarantees in dynamic environments compatible with many existing control algorithms. Provides software implementation with detailed documentation for non-experts, which currently does not exist in the community.
pCBF is a MOOS-IvP app for maintaining safety of autonomous vehicles using Control Barrier Functions (CBFs). Among other things, pCBF can be used for
- Preventing inter-agent or agent-obstacle collisions in multi-agent, multi-obstacle settings
- Preventing agents from entering "keep-out" regions
- Keeping agents inside "stay-in" regions
Note: pCBF is still under active development. Expect bugs and incomplete features.