One Obstacle Avoidance Method-Artificial Potential Field (Part 1)


One Obstacle Avoidance Method-Artificial Potential Field (Part 1)

By: Mei Jian Hong (PhD Student)




Artificial Potential Field (APF) is a reactive method for robot path planning, which was proposed by Khatib in 1986. It uses a position related potential function to achieve obstacles avoidance, and it makes the robot flexibly cope with the changing environment.

Assuming that the there are obstacles standing between start and target in robot motion space, the robot is required to plan a free collision path from start to target. The idea of APF method mainly consists of virtual force vectors, where obstacles repulsive forces and target attractive forces work on the robot simultaneously. Thus the motion of robot is determined by the resultant of forces, as shown in Figure 1.


6a june15

Figure 1. Illustration of virtual repulsive and attractive forces


The repulsive forces induced by obstacles are represented by repulsive potential function which is related to the distances from robot to obstacles. The potential function can be expressed as,

6b june15

6c june15


  1. O. Khatib, Real-time Obstacle Avoidance for Manipulators and Mobile Robots. International Journal of Robotics Research, 1986, 5(1): 90-98.