While the Large Hadron Collider is shut down for repairs, you might be feeling pessimistic about grand scientific experiments. But that's the cool thing about science - even when everything goes horribly wrong, we still learn something. Sometimes, what we learn from failure is more important than what we'd have gained from a success. Here are five scientific experiments that didn't go as planned, and we're all better off for them.Penicillin - Alexander Fleming was studying bacteria in his own messy way, with no intention of discovering the 20th century's most vital antibiotic. Indeed, his lab sounds like something out of a sci-fi/horror movie, with bacteria and random fungus growing everywhere. Some of the accidental fungus had been tossed away, but looking more closely, Fleming noticed that bacteria wouldn't grow near some of the stuff. It took the work of others to refine and mass produce the extracted antibiotic substance, but if Fleming kept a neater shop, we may never have found it to begin with. The Aether Wind - In the 19th century, physicists were stumped by the nature of light. It seemed to behave like a wave, so there had to be some substance in space for it to move through. They dubbed this hypothetical intergalactic substance "aether." It was theorized that the motion of the Earth through space, relative to the motionless aether, would subtly alter the speed of light depending on where in its orbit Earth was and what direction you were facing. This was called the "aether wind" effect. Polish-American scientist Albert Michelson (Polska represent!) designed an interferometer that could precisely measure the speed of light and thus detect this wind effect. After several tries and refinements to make his device incredibly accurate, no change in the speed of light was detected. Michelson, along with pretty much every other physicist at the time, was stunned. No aether? WTF? Rocketry - No scientific failure is perhaps as spectacular as that of a rocket exploding on the launch pad, like the Vanguard rocket expiring in the 1957 photo above. The rockets that have died in the name of science number perhaps in the thousands, yet they did not die in vain. NASA and other government space agencies can put people and payloads into space with astonishing consistency (private rocketry is still catching up), giving companies the confidence to send aloft hugely expensive satellites and ambitious scientific equipment. Our world would be very different if we hadn't learned so much from all those shattered rockets. Biosphere 2 - We built a big dome and let a bunch of people live in it (none of them were Pauly Shore) to see if they could sustain themselves solely on the air, water and food produced by the plants inside. They couldn't. The overriding element of the Biosphere 2 experience for most participants was "hunger." But when we build a colony on the moon or Mars or somewhere even more interesting, we will build on the lessons learned via Biosphere 2's rampant pizza cravings. Nuclear Fusion - Is it a pipe dream or a holy grail? Either way, each failed experiment brings us one step closer to deciding that fusion is not worth pursuing any longer/going to provide us with so much energy we'll be giving it away. There have been lots of failed fusion experiments, but one of the coolest happened in 2002, when scientists sent incredibly strong sound waves through acetone. This created bubbles that expanded, then imploded at very high temperatures. It was hoped that the temperatures and pressures would be high enough to foster a fusion-friendly environment, but they fell a few million degrees short. Still, it hasn't dampened our enthusiasm one bit. Honorable mention goes to Chernobyl. It was an ill-advised emergency shut-down experiment that caused that catastrophic meltdown and explosion there. Image by: NASA.