Please try this once even if it sounds weird!
A lot of engineering consists of doing research, design, analysis or testing and then writing a report. Next time, try writing as much of the report as possible before doing anything else.
Writing the introduction will force you to clarify what you are trying to achieve and why. Documenting your assumptions will force you to think about them explicitly. Documenting a planned test method often highlights flaws in the plan.
Writing the headings for the rest of the report will force you to think about the scope of the work you are about to do. Is all the work you were planning actually necessary? Are there extra tasks which you need to do?
If you are doing research, design, analysis or testing and not delivering a report, then please consider writing your work up as a Technical Note and applying the same process. I find that the small extra investment in time usually pays back immediately by forcing me to think about my plan more clearly before I start, never mind the long-term advantages of having my work clearly documented.