NERO is a result of an academic research project in artificial intelligence, based on the rtNEAT algorithm. It is also a platform for future research on intelligent agent technology. The NERO project is run by the Neural Networks Group of the Department of Computer Sciences at the University of Texas at Austin. To learn more about NERO, check the About page and the illustrative videos. To try out NERO, download the game binary and run the tutorial. To use NERO in your research or teaching, contact NERO team.
http://www.nerogame.org/

This is a great site. They use the Torque Game Engine for their development, but the code is free (this is a University class). All of the code is in classes and the logic is easily converted to lite-c. Also the learning curve here is awesome. I keep bugging them to do an online class, but to no avail.

To get the source code I simply contacted the NERO team and they emailed me a link to the download.
Have fun.

Quoting ATOMIX.

...

Are classes available in Lite-C also?

There is an interesting Paper on the basic Mechanics they used in their approach.

http://nn.cs.utexas.edu/pub-view.php?PubID=97113

I had the same question and it appears that the t7 programming interface with A7 is a group of classes. Also the manual states that you can use Windows classes by using COM. I haven't done either one yet and am still waiting for examples for both of the the above.

I like it a lot. But it is pretty difficult to read the code. It would be a lot better to have something like a tutorial...

Thats why there is a white paper...

What do you mean?

Read the white paper in the link I posted,
this explains the structure of the approach,
and is more helpful then the raw C++ code.