009 The expensive mistake in every overbuilt prototype

Speaker A: I was working on my garden tractor and it got me thinking about a video I made a little while ago where I suggested that when you do some testing on your prototypes, you can actually find some areas that are overbuilt and you can take some material out of that overbuilt part and then reapply it to some other areas of the product that may not be strong enough. In this video, I was accused of encouraging planned obsolescence in designs by the idea that if you make a part weaker than then you're increasing the likelihood that the entire product is going to break. I disagreed with this comment. And about a week later I was having a discussion on LinkedIn where I made a post about the difference between overbuilding and over engineering. One of the comments I received was really interesting and it made me think back to this video. And the guy said, anyone can build a bridge, but to design a bridge that barely stands up, you need a good engineer. And this isn't about trying to make your products fail, it's about, it's about knowing at what point your product will fail so you can design beyond those limitations. And this is why we have professional engineers designing bridges, because these are things that can't be prototyped and they can't fail. I personally am, um, focusing on product design and there's very few products that actually have to be signed off on by a professional engineer. Now we can use computer simulations and do things like finite element analysis to try to get an idea of what the maximum stress that a product can take. However, keep in mind that you still have to model it correctly. You have to know what stresses this product will see in real life. And oftentimes we don't really know as the saying for computers goes garbage in, garbage out, and you make one little mistake there and your model could not represent real life at all. So what this ultimately means is when most people design a product, whether you're an engineer or not, it's, you're taking a lot of guesses on where it's going to fail and where it's not. And as a result, a, uh, common practice tends to be to over build whatever you're designing. And I do this a lot, especially when I'm designing and building machines that made in low production. The problem with overbuilding to try to prevent failures is you very rarely know where the failures are going to be. And culturally overbuilding is seen as a sign of quality. No one's going to follow you if you overbuild something, you make things thicker, stronger, more Robust and then wear it as a badge of honor saying that you did quality. However, overbuilding just for the sake of overbuilding is a very lazy way of doing things. You end up with something that's heavier, more expensive, harder to sell. Overbuilding also gives you a false sense of confidence because as I said before, we're designing around what we think the expected failures are going to be. And by overbuilding one part you could actually be stressing another part because it now needs to handle a heavier part that that might put more load on it than you originally intended. So what's the correct solution? The correct solution is properly engineer it. And I know sometimes it's hard to pay for all the engineering that needs to go into a product. And this is especially the case for startups, early prototypes or something that very few production units are going to actually be made for. So we know that overbuilding can save your ass if you don't have time to do it right. We know that thorough engineering can help you optimize the design if you have the time and money. However, simplicity beats everything because keep in mind, you can't break apart if it doesn't exist. So for those parts that are left that you have to do something about, you want to get testing as quickly as possible. Build your prototype, build it simple, build it cheap, and on top of all, under build your prototype. We're not trying to prove that it works, we're trying to prove every way in which it can break. And every time you get a failure, you beef it up just a little bit until you solve that problem. And, and when you have all the problems solved and you know where your failure points are, that's when you can beef it up and make it stronger than it originally was intended. You're going to design it, you're going to build it, you're going to test it, you're going to break it, and then you're going to do that over and over and over again until there's nothing else you can learn from that prototype. Hold off on refining your prototype. The last thing you want to do is be refining something that's not there yet. You're trying to look for aspects of the design that may not be needed or anything you can simplify or, or maybe an original assumption you had about it that just isn't true at all. Refining your design before it's ready is like trying to sand down something with 1200 grit sandpaper when it's still full of 80 grit scratches. You're just doing extra work at this Point that you're then going to have to redo from scratch later. So, getting back to my project here, this tractor has a bad fuel pump, and it's a very cheap part. I think I paid like $4 for it, and it's easy to replace, so that's no big deal. However, what is a much bigger issue is this engine has what's called an automatic compression release on it, which is a feature on the camshaft that holds the exhaust valves open while the engine starts. This engine has high enough compression that it's very difficult for the starter to turn it with those valves closed. Once the engine starts, then that feature can release the exhaust valves so they can close and the engine can run properly. Now, we could have beefed up any part in this tractor. The fact that this engine is going means that none of that matters. We could have beefed up the metal that the body of the tractor is made of weight. We could have put a heavier mower deck on it. We could have increased the strength of every part on this to try to make it a better tractor. And all those parts would have last until the iron moths reclaimed it to the earth. However, the fact that the engine goes bad, I don't want to rebuild this. I'm not going to take it apart. I'm not going to put a new camshaft in it. I'm not going to go out and spend $1,000 to put a new engine in a garden tractor that I paid $150 for. And is that ridiculous? No. I can go on Facebook and I can buy a working garden tractor with a working eng for $400. So you need to test. You need to find your failure points and don't just overbuild other parts and get a false sense of confidence that what you have is great, allowing you to overlook where your real issues are at. So I'm just going to put a new battery in this and deal with the hard starting for as long as I can until no longer starts. The one mindset shift that is going to make any product you design better is going to testing trying to break your prototype, not hoping you won't break it. Plan to break it 10 times. Take that as a baseline. So every time you break it, say, excellent, we have found one failure. And then you improve that and you try it again, and you break it again and you find another failure. Every failure is a learning experience, and every one of those failures makes your final product better quality. Keep your prototypes cheap. They shouldn't be pretty, they should be ugly. That's the whole point of them. When you break it, try to fix it and reuse it, unless it's a bad idea and you need to go in a different direction. So make your new prototype with your new idea, but don't throw out your old prototype, because sometimes your genius new idea has a major flaw that you never thought about. So you want to have all the different versions, and you'll probably end up using something from the old version, something from the new version, and, and put it together and come up with a third idea that works better than any of it. And don't ever just test your prototypes for the expected use. Try the edge cases, put it in dirt, overheat it over, stress it, give it to your second favorite child to use as a hammer. The perfect prototype and the perfect testing will let you know what the failure point of every single piece in your design is. And that's unrealistic, but that is where your goal is. So that's why I said I underbuild it. And try to break as many parts as you can and then find out at what point that those parts don't break anymore. If your goal is to make every part twice as strong as it needs to be, you're going to have no idea what twice as strong as it needs to be is if you can't break it in the first place. A part of a design that's skated through testing without having any problems could be five times stronger than it needs to be. Or it could have been just barely strong enough to make it through all the testing. If you really want to understand the mindset of of the different ways to design a product, check out my other video that I made that covers the difference between overbuilding and over engineering. M.