I lean towards agreeing with eykanal and disagreeing with Sylvain.
A PhD project is inherently a somewhat fungible plan—what you do depends on the results that you've obtained. And obviously equipment failures, unexpected obstacles and delays, and other unforeseen circumstances are a natural part of research. So I wouldn't be too concerned about failing, as a graduate student, to reach the long-term destination of your research project.
In particular, with my own students, I try to sketch out as little as possible the actual outline of the projects they are going to pursue. That way, there is much more flexibility in the future development of their projects, as I plan to tailor them based on the students' expertise and interests. (Moreover, I would suggest that if you know everything you need to do to reach your goal, and you accomplish exactly that, you haven't done any research at all!)
That said, there is the issue of meeting short-term goals as well as long-term goals. You shouldn't promise your advisor something will be done in 1 to 2 weeks if you don't intend on having it ready for 2 months. (I would also argue the reverse is a dangerous situation, too, because you could fall into the "competency trap," whereby the advisor thinks you really know what you're doing, and continues to expect you to know what you're doing for the remainder of your time in the advisor's group!)