Steering Behaviors Part 1: Seek

This tutorial shows an HTML5 implementation of the Seek Steering Behavior in which a vehicle (seeker) moves in a realistic manner toward a target using the seek algorithm.

The Steering Behaviors are artificial intelligence algorithms which allow autonomous characters to move in a realistic manner. They are developed by Craig Reynolds. More about the theory of steering behaviors algorithms you can find in his papers.

The first steering behavior which will be discussed here is the seek algorithm. Its presentation is divided into next three sections:

  1. a simple live demo programmed in HTML5 using Phaser framework
  2. images with used math vectors and a pseudo-code explanation
  3. full source code listing which is also available for download on Github


Live Demo

The green arrow (seeker) seeks the blue circle (target) and moves toward it. Use your mouse to move the blue circle (target) on some other location.



Vectors and pseudo-code

Steps 1-3: Calculating the desired velocity


  • 1. Vector (desired velocity) = Position (target) – Position (vehicle)
  • 2. Normalize Vector (desired velocity)
  • 3. Scale Vector (desired velocity) to the maximum speed


Steps 4-5: Calculating the steering force


  • 4. Vector (steering force) = Vector (desired velocity) – Vector (current velocity)
  • 5. Limit the magnitude of Vector (steering force) to the maximum force


Steps 6-7: Calculating the vehicle’s new velocity


  • 6. Vector (new velocity) = Vector (current velocity) + Vector (steering force)
  • 7. Limit the magnitude of Vector (new velocity) to the maximum speed



The source code



2 thoughts on “Steering Behaviors Part 1: Seek

Leave a Reply

Your email address will not be published. Required fields are marked *