Reactive Navigation of the Ackermann Steering Robot in Unknown Environments

Mobile robots are now widely used in numerous
real-world applications that have complex navigation
requirements, especially in environments used by humans. This
requires highly accurate navigation that can be performed in real-
time. In this paper, a method for generating a smooth motion of
nonholonomic mobile robots is proposed. It enables robots to move
optimally toward the desired goal and allows fast path replanning
when encountering unknown or dynamic obstacles. The method
generates smooth, collision-free trajectories based on clothoids,
ensuring high computational efficiency and suitability for real-
time path planning. By applying a smoothing algorithm, the
proposed method improves the robot's efficiency in terms of travel
time and trajectory length from start to goal, as demonstrated by
a comparison with model predictive control.