The Dynamic Window Approach to Collision Avoidance presents a method for reactive collision avoidance in mobile robots with synchro-drives. The approach is derived directly from the robot's motion dynamics, making it suitable for high-speed operation. It reduces the search space to the dynamic window, which consists of velocities reachable within a short time interval. The method considers only admissible velocities that allow the robot to stop safely. The optimal velocity is chosen by maximizing an objective function that includes progress towards a goal, forward velocity, and distance to obstacles. The approach has been tested on the RHINO robot, achieving safe operation at speeds up to 95 cm/sec in various environments. The method is robust and efficient, incorporating the robot's dynamics and constraints. It has been implemented using ultrasonic sensors, cameras, and infrared detectors. The dynamic window approach is particularly useful for robots with limited motor torques and high speeds. The method has been shown to effectively avoid collisions in cluttered and dynamic environments. The approach is part of a larger system that includes global path planning and computer vision. The method is well-suited for proximity sensors like ultrasonic transducers and laser range-finders. The approach has been tested in various environments, including the AAAI '94 mobile robot competition, where it demonstrated effective collision avoidance. The method is robust and efficient, incorporating the robot's dynamics and constraints. The approach is part of a larger system that includes global path planning and computer vision. The method is well-suited for proximity sensors like ultrasonic transducers and laser range-finders. The approach has been tested in various environments, including the AAAI '94 mobile robot competition, where it demonstrated effective collision avoidance.The Dynamic Window Approach to Collision Avoidance presents a method for reactive collision avoidance in mobile robots with synchro-drives. The approach is derived directly from the robot's motion dynamics, making it suitable for high-speed operation. It reduces the search space to the dynamic window, which consists of velocities reachable within a short time interval. The method considers only admissible velocities that allow the robot to stop safely. The optimal velocity is chosen by maximizing an objective function that includes progress towards a goal, forward velocity, and distance to obstacles. The approach has been tested on the RHINO robot, achieving safe operation at speeds up to 95 cm/sec in various environments. The method is robust and efficient, incorporating the robot's dynamics and constraints. It has been implemented using ultrasonic sensors, cameras, and infrared detectors. The dynamic window approach is particularly useful for robots with limited motor torques and high speeds. The method has been shown to effectively avoid collisions in cluttered and dynamic environments. The approach is part of a larger system that includes global path planning and computer vision. The method is well-suited for proximity sensors like ultrasonic transducers and laser range-finders. The approach has been tested in various environments, including the AAAI '94 mobile robot competition, where it demonstrated effective collision avoidance. The method is robust and efficient, incorporating the robot's dynamics and constraints. The approach is part of a larger system that includes global path planning and computer vision. The method is well-suited for proximity sensors like ultrasonic transducers and laser range-finders. The approach has been tested in various environments, including the AAAI '94 mobile robot competition, where it demonstrated effective collision avoidance.