[intro plays]

Sabrina: Hey future engineers, I see you successfully made it back so we can find out why it's so important to fail when we're testing a solution to a problem. Last time, we discovered that when a solution to a problem doesn't work anymore, it's called a failure point. But, knowing the definition of a failure point isn't enough. Engineers have to be able to spot failure points, and then try to fix their solution,  so those failure points won't be a problem anymore. So, we've gotta ask: how do engineers find and fix failure points in a solution they're trying to develop?

[text: Big Question]

Well, for one thing, when they're testing a solution, engineers don't just jump in randomly, they make a plan. That is, an organized series of steps used to accomplish a goal, which in this case is a successful and safe solution. Now at this point, you won't be too surprised to find out that this plan will include isolating a variable and then doing a series of trials or tests, changing that variable each time until you find a solution that works. That's what we did at the bowling alley, remember? We isolated the angle of the ball ramp as a variable and then changed it until we got a strike. Success!

But engineers don't stop there, they try to find any and all failure points that might come up. So, to see how we can sniff out any possible failure points, let's go back to our little wooden bridge in the forest.

[text: Investigation]

Remember when we needed that bridge to cross a stream? We decided that the bridge could take us safely across the water, but not us and our bulldozer. But, just guessing that the bridge can't support a bulldozer isn't good enough, engineers would need to make a plan to test the bridge, and see if they can find failure points first. One way that engineers test a solution in its early stages is to build a small model of it first.

So, say we have the materials we need to make a model of a bridge. Once we're done building it, we need to plan to test whether our bridge can hold a certain amount of weight, or not. So let's keep track of how much weight the bridge holds as we pile weights on its deck. We'll keep adding weights until we find the amount of mass that makes the bridge collapse. And down it goes!

Now, that amount of mass was less than our goal weight, which we'd need our model to support if our real-life bridge was to hold up our bulldozer, so guess what? We've just identified a failure point. We have a specific weight at which the bridge is no longer a good solution. But, before we dump our bridge idea and start again from scratch, let's see what happens when we change the existing design. Let's build the bridge again, and this time add extra support. Now, let's follow the same plan as before and keep adding weights to the model to see how it holds up. In this new model, we see that the bridge holds our goal weight. In fact, the bridge holds even more than that. So far it looks like our design could be a successful solution. But, there are more possible failure points than not being able to support enough weight.

Remember when we talked about the Tacoma Narrows Bridge? It was a perfectly fine solution to the problem of getting people across Puget Sound, at least until the wind picked up and hit the bridge at a certain angle. Then the bridge became good ol' Galloping Gertie, twisting in the wind until it collapsed. So in that case, the speed and angle of the wind made up that bridge's failure point. So even though we've determined that our bridge can carry enough mass, we should still pick out and test other variables and try to find other failure points to make sure we've built the safest bridge possible. So, let's review some of the steps that engineers
go through when they're trying to find and fix failure points.

[text: Conclusion]

To figure out whether a solution is successful, they make a plan, or an organized series of steps to accomplish our goal. And part of this plan includes isolating a variable and then changing that one variable in a series of trials or tests. And once the solution fails one of those trials, they've found a failure point. So, if you actually plan to fail then you're not really failing, you're making your solution better.

[outro plays]