I identify myself as a human agent. It is time to think about oncoming senior research and due to small experience in gamedev(as well as in AI field), some questions are raised. What are the most suitable approaches to implement real-time simple AI agent in an action game? I've heard something about cognitive architecture like ACT-R.

By design, entity's AI can have several mutually exclusive states. enter image description here

This is an existing AI of game, which has states, events and schedules. However, the code is complicated and not flexible. Also, it does not use any cognitive architecture, which I consider as a drawback. https://www.youtube.com/watch?v=9jO-P3kXlCI

Please, using your experience suggest any modern techniques, which can copy such behaviour as in image or video.

Thank you for your perception.

  • $\begingroup$ You're on the right track though here;concerning Artificial Intelligence community principles,we don't recommend resources.also your question should be reasonably scoped. If you can imagine an entire book that answers your question, you’re asking too much. $\endgroup$ – quintumnia Feb 24 '17 at 8:09
  • $\begingroup$ Thank you. Now I removed useless text. Can you read it again? $\endgroup$ – null Feb 24 '17 at 10:17
  • $\begingroup$ I think it would help if you could be a little more clear about what exactly you're asking. Maybe describe the game and what your agent should do, in text form? A lot of people will probably be put off by the need to click through to a video in order to properly evaluate the question. $\endgroup$ – mindcrime Feb 26 '17 at 9:52

About 15 years ago, John Laird's group at Michigan used the Soar rule-based architecture to play several FPS games effectively (Quake II, Descent III):


Here's Laird's overview article from 'Computer':



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.