All right, in this lecture, we're going to take a look at Pugh matrices, which are a tool that you can use when you're doing design, prototyping, really, any place during your design process that requires some decision based on ranking criteria. This is a common tool that comes up in lean manufacturing. It's one of the more effective ways to get a team to focus on trying to make sure you're making a decision based on the right criteria. It's part of a methodology called set-based design, which we'll also talk about a little bit here. So consider a situation where you're working on a protocol assessment. You're designing, say, an IoT style application. And the folks on the team know a bunch of different IoT application protocols like MQTT, or CoAP, or what have you. And the customer really didn't specify a particular protocol, and they really don't know what they need from your application. How are you going to do this? How are you going to pick? You might have people on a team that have used one of these before. You might have products that have used some of these before. But usually, there's gotta be a little more reason than that to make a decision on which way to go for a technical solution. The possible criteria for selecting a protocol are pretty immense. You could spend a lot of time considering a lot of different criteria, whether it's the size of the messages that are being sent. What kind of tools are available? What sort of international use is there for the particular protocol? What's its range? What's its data rate, etc? Could you balance all this? Or should you even try? I think that's really the question that we're going to answer with cue matrices. Another typical place where criteria based decision for IoT devices shows up is when you're selecting components or sensors. This is another problem area, because there are potentially so many competing criteria to take a look at. Whether it's the functionality of the sensor, its data rate, how big it is, how much power it uses, what its ranges, etc. All those things have to be considered. And while you might be lucky and have one or two that bubble up immediately as the most important issues. In most cases, you're going to be balancing several criteria, and you have to think about how you're going to select those. In set-based design, the goal really is to try to trim away the design sets that don't particularly fit an application early on. And you will do this by spending some time upfront learning, identifying risks and mitigation, and doing alternative assessments. Just enough to make sure that the decision space that you're looking at doesn't include anything that's really an inferior alternative that is pretty easy to to eliminate. It's difficult in all design processes to avoid bias, to avoid moving towards a favorite solution. But the Pugh matrix, which is part of set-based design, does help us stay away from things that might be a personal favorite for some team member. Set-based design is an alternative to point-based design, where in a point-based design, you're kind of going down a single path. You create a design. You analyze it. You critique it. That may involve some other teams like marketing or manufacturing. It may involve proof of concept work, or prototyping, or research, what have you. You look at how that design came out, and then you make modifications and repeat. The problem with that is, again, you can easily get yourself down a path with a single flow like that, where you run into some critical issue that you hadn't considered, and you find yourself having to loop back and start over. And it costs time, it costs money, and it delays your product. In a set-based design, you're going to spend, again, some time upfront trying to eliminate what they call dominated alternatives, where those particular alternatives don't meet some of the requirements from all the different sources that are coming into the design. And you're going to do the best you can to quickly assess or experiment to trim away the most obvious ones from the design set. And then focus in on studying the design alternatives that make the most sense for what you have. So one of the ways to get to that point is to use a Pugh matrix. Pugh matrices have been around for a while. It's a lean engineering manufacturing tool. It also is found in Six Sigma quality processes. It may also be known as you might see it as a criteria based matrix. What you're really doing in a Pugh matrix is essentially, you're identifying a set of criteria that you're going to use to assess a given set of solutions. Hopefully, you'll also have a current baseline solution that you can compare against. Although, that's not necessarily required. But once you have those, you can start ranking the solutions, giving them some sort of scoring. And then you can sort of add them up and decide whether or not you're making a good decision. The quick example here shows buying a car where you're looking at things like fuel consumption, how many doors it has, etc. And you're comparing the alternatives you're looking at to the baseline car that you already have. Hopefully you can kind of see that even in this simple example, by concentrating on their criteria that are most important to hear, and then scoring your different alternatives, you should be able to make a pretty good decision, and use it to make a pretty quick decision. The Pugh matrix process in general is not a difficult one. You do want to include the right people from the design team, have some sort of a template that you're going to use, and then walk through an enter values for a baseline design if, indeed, you have one. If you don't, then you'll have to be comparing the different alternatives to each other, which is also possible. But the baseline makes it a little easier to think about what's good, or what's better, or what's worse than another design. You then go through, and you list what those alternative approaches are that you're considering. And then go through and look at which of the criteria the team believes are the key criteria to be used in this assessment. Ideally, you're going to weight those criteria, so that the most important criteria have the biggest impact on scoring. Then you go through, and you've ranked the concepts against each other, and against the baseline. So you can kind of see how this works. And once you sum up those scores, you should be able to see that select alternatives are scoring better than others. And this might cause the team to cycle on this just a little. If the scoring doesn't come out the way you expect, you might start looking harder at how you're weighting criteria, or how you're assessing product, and that's all good. It's part of the conversation that needs to happen during this process. So imagine again if we were looking at trying to do a Pugh matrix for protocol selection. Hopefully, again, if we had a baseline protocol, we might compare against it. If not, we could choose one of the alternatives as a baseline. We would go through and list all the different protocol options that we're considering. We would look at what the key design criteria are, and we would weight those criteria. Studying different ways, you can just use straight 1 through N for how many you have. You might weight them one, two or three. It really doesn't matter too much, because, again, as you score, you'll start to see which of the alternatives are really meeting the most important criteria. Then you have to go through the score the possible alternatives, or options that you're looking at. Sometimes, we do this with a set of minuses and pluses compared to baseline. You can also do this with a 1/3/9, or what have you, to make your choices a little clearer. And then look at the results. The most important thing in a Pugh matrix which is easy to fall into, is not to weight things that are favorite choices unfairly. You really want to give the design team the instructions that they want to focus on trying to do this as Independently, and with as much consideration as possible, to make sure that the choice that you're making is a good one. So that's pretty much it. A Pugh matrix analysis can provide a nice way to focus a team in on a criteria based choice. And it helps trim away solutions that don't fit without spending too much time on them. Again, there is this danger of scoring favorite choices. But again, usually, if you're working with a design team, talking through those choices is going to make it pretty clear when people feel like they have some skill in the game, and that they're really biased towards something. And again, it's good to have those conversations, and talk about why people are making choices. It's also a really powerful way to present your choices to customers or stakeholders to be able to clearly explain why you're making a decision, and what criteria are you using? And again, it does provide a good team building approach to getting the team into a design space quickly, and focusing on the things that matter. So I hope this is another tool that you can add to your toolbox that might help you through your own design processes. All right, onward. Thank you.
