Abstract: As the complexity of robot system working environments continues to increase and the demand for real-time performance gradually escalates, the safety avoidance capability of robots is facing new challenges. Control barrier functions, as the safety method based on controllers, are getting new development opportunities in robot safety control systems. This paper investigates and analyzes control barrier functions and optimization controllers based on quadratic programming, summarizes the obstacle avoidance problems of robots in known and unknown environments, and provides an overview of strategies for synthesizing control barrier functions from the theories of Gaussian processes and reinforcement learning. Finally, it systematically discusses the key issues that need to be focused on in the future for safe control of ground collaborative robots based on control barrier functions, providing inspiration and references for future theoretical research and technical applications of control barrier functions.