You start out with a simple question, you come up with a plan to answer it, you find out that your plan actually answers a different question, you realise that the new question is far less interesting than the original question, you run out of time and publish the answer to a third question that really was not asked. You never know what you're gonna get.

What should have been done in your thesis:
How many rubber duckies fit into the box?
What you wanted to do in the beginning:
Laser scan
3D-reconstruction
Monte-Carlo-simulation
Optimal object placement
What you actually did in the end:
half-assed laser scan
Polygone pile
Segmentation fault
Just putting the duckies in the box with a mechanical press
What you wrote in your thesis:
When the brute-force-approach is chosen, the rubber duckies break.
Future Work:
To find out the optimal number of duckies, one might consider using a Monte-Carlo-simulation...