Feedback

This is challenging and open-ended topic. The simulation of responses is clearly very challenging in general, and there are technical challenges at both ends of the manual-automated behaviour spectrum. If the modeller is to act as driver, the challenge is 'how to convey the situation effectively?' - if the model is fully automated, then 'how to simulate non-standard responses?'. Normally in EM we would aim to move from the manual to the automated, but this doesn't seem to be such an attractive strategy in your context. I like the way that you are setting out to address this problem by using an LSD analysis.

The discussion you have developed so far is strong on the way in which the modelling principles are being applied, but not as yet so well-oriented to the concept of 'writing a paper about EM'. It may well be that you can present the kind of discussion you've ventured so far in the context of an evaluation of the potential for EM, and a critique of the principles and tools for model-building as so far conceived. (For instance, I've already acknowledged that your chosen problem doesn't seem to fit the 'from manual to automated' paradigm well!)

It may be possible for you to make use of a USB steering wheel device if you would like to add a more realistic element to the car driving mechanism.

References are rather thin at this point! It may be that few models have been directly referenced in your development to date, but it's probably a good thing to review ones that could be useful in connection with possible extensions beyond the scope of the assignment (such as carparkingsimMcHale2003). It would be good to find a few references to simulators that are used in training police drivers etc, even though these are also well beyond the scope of what can be emulated.

Good to try to improve / simplify the English in places: e.g. 'as further definitions ... would be very large in quantity' and 'will be defined as repetitive i.e. ...' are rather clumsy expressions.