Possible Research Question

This week I've come up a possible research question to base my honours project around. So here it is:-

"Is it possible to implement cognitively modelled,

self animating non-player characters

into a computer game on a mass scale and if so

what is the best method for achieving this?"

To answer his question I plan to research the different methods of producing cognitive models so as to be able to compare and contrast these methods effectively. From initial research it seems that 3 main areas have come to light, these are a fairly simple rules based system, a fuzzy logic rules based system and a system known as Cognitive Modelling Language (CML). I openly admit that I need to do far more research into each of these areas to see if each one of these is feasible but they do seem to be the most likely solutions from my initial reading.

Once the research has been completed I plan to try to implement each of these techniques into a game engine and test performance. The engine I will most likely use will be a version of the OGRE engine adapted for use as a game engine. Using OGRE this way takes away alot of the burden of creating an engine from scratch and has the added bonus of vast online support meaning I can rest easy knowing that the problems that I encounter along the way will be entirely relevant.

The main focus of my question is the idea of implementing these solutions on a mass scale, so "performance" does not necessarily refer to traditional methods of testing efficiency. By performance I mean literally how many models animating themselves in this manner can I put into the game environment whilst maintaining a high level of believability and realism to their actions before the game drops below an acceptable frame rate.

While this is only a draft question it seems to me to have the potential to be very interesting and above all produce good results.