Integrated Obstacle Avoidance and Path Following Through a Feedback Control Law

The article proposes a novel approach to path following in the presence of obstacles with unique characteristics. First, the approach proposes an integrated method for obstacle avoidance and path following based on a single feedback control law, which produces commands to actuators directly executable by a robot with unicycle kinematics. Second, the approach offers a new solution to the well-known dilemma that one has to face when dealing with multiple sensor readings, i.e., whether it is better, to summarize a huge amount of sensor data, to consider only the closest sensor reading, to consider all sensor readings separately to compute the resulting force vector, or to build a local map. The approach requires very little memory and computational resources, thus being implementable even on simpler robots moving in unknown environments.