Bot Navigation: Design Philosophy

By ai-depot | June 30, 2002

Conclusion

In this essay, we discussed the importance of spatial awareness and terrain knowledge. Having shown that current bot navigation is far from perfect, the human approach to this problem was analysed. That allowed us to define a set of requirements for a system that is capable of handling complex environments in a flexible fashion. A brief outline of the system’s framework has also been established.

Our task from here consists of the following parts:

  • Raw obstacle avoidance will be covered first, and a few experiments with neural networks will allow us to determine a robust model.
  • An automated waypoint system will be described, based on human navigation.
  • Neural networks will then be trained to explore a level, thereby automatically creating the waypoints
  • The dynamic path-finding algorithm will be designed.
  • The neural networks will then be trained to follow a given path.
  • Final touches will then be performed to the system to allow a more human-like behaviour.

Stay tuned for the unfolding of this series.

Written by Alex J. Champandard.

Pages: 1 2 3 4 5 6

Tags: none
Category: essay |

One Response to “Bot Navigation: Design Philosophy”

  1. alhnuf Says:
    October 7th, 2007 at 8:44 am

    thanks .. .good

Comments