Originally written as an explanation for my qualitative research methods students.
The discussion section of your project is where you answer the question: "so what?" This typically comes at the end, because you are discussing that question in relation to the results of your project, not the problem statement like you did at the beginning. Here's the difference.
The introduction comes before you talk about the results, and tells us what problem you're trying to solve (or what question you're trying to address) and why it is important. For instance, "we should find out what kinds of things make Superman weak, because if he's going to keep saving the world, we need to know what might prevent him from doing so" would be an introduction that explains the motivation behind the research question of "what effect do various ore forms of radioactive elements have on the strength and flight abilities of Superman?"
Now, suppose you find that among the elements you tested, only kryptonite has a measurable effect on Superman's strength and flight abilities, and that it strips him of his powers and makes him physically ill. Those are your results -- see how they're a direct answer to the question you asked earlier?
The discussion comes after your results. Now that we know that kryptonite weakens Superman... so what? What difference does this specific answer make to how we operate? Well, maybe we want to preserve Superman's superpowers, so this means need to make sure Superman doesn't come in contact with Kryptonite... perhaps we should make it illegal to possess the substance. Or maybe we want to give a subset of people access to Kryptonite so that they can take Superman down in case he goes evil.
That's a discussion. Note that the discussion depends on the results you got -- if the results change, the discussion changes. For instance, if you found that no ore forms of radioactive elements had measurable impacts on Superman's abilities, you wouldn't recommend outlawing Kryptonite because it'd make no difference in this case. If you found out that all ore forms of radioactive elements came close to killing Superman, but only on alternate Tuesdays, and only if he ingests them, you might talk about creating a Superman Radioactivity Food Scanner that only operated once a week to save resources. You get the idea.
The discussion question of "so what?" is not a question you need to have an answer to until the end, but you should know at every point in time that you will need to answer it... in other words, it's something to constantly keep in mind, and you'll find it along the way as you develop your project. The introduction and discussion are both important places where you tell us why your work matters -- but the introduction is where you tell us why the question is important, and the discussion is where you tell us why the answer is.