• control
• danger
• planning
• risk
• robot

The goal of this study is finding a mathematical definition of risk that goes beyond the simple distance to the obstacle, combining a realistic representation of doubt about a possible collision without taking cover with exaggerated confidence margins. In this way, the robot can maintain high safety standards without compromising efficiency. The hazard function must be minimizable in a simple way directly in the robot's configuration space so that a control law consistent with its kinematic structure can be derived. The control algorithm must also be easily adaptable to different scenarios and robots and stackable to all kinds of tasks.