Thinking before acting: AI planning in games

The more complex and open ended your game is, the harder it becomes to consider every possibility that your AI will need to handle.  AI Planners alleviate the need to consider everything.  Given a goal, they can figure out a series of actions that will achieve it. This can empower your agents to solve their own problems, and can help provide more robust, and emergent AI.

This talk introduces the high level concepts of how common planners work, how to use them effectively, and when they are worthwhile.  Finally, some extensions and optimizations that can be used to further improve your AI will be covered.

Game Development
Location: Ottawa School of Art Date: April 25, 2016 Time: 11:15 am - 12:00 pm E. Hahn Pic Evan Hahn