I think if the advisor and the student stipulate unambiguous targets in advance, then the roadmap for both of them should be clear. Nonetheless, there might emerge exogenous impact factors (private situation, lost novelty) that require profound amendments.

Dear Mohammad, 

In Sweden their is a special and helpful training for supervisors of PhD-thesis to understand their students. I have observed that if your student admires you in a very outstanding way in the beginning of the process be aware he/she will hate you at some point. Sometimes you as the grown-up and hopefully wiser one has to back off, even apologize, - if necessary - to get the process going. At some instant the student starts to feel more clever than you and consider you as almost stupid---try then to remember how you felt towards your supervisor, this makes it easier to digest! Good luck!