This video walks through taking the information from the first video, the preparation for FMEA, and applying that to the FMEA form to allow you to prioritize your actions moving forward to improve your design.

In this video we'll walk through the results of the preparation, that's the design tree, the function tree, and the failure list.

And then we'll look at the FMEA form and how those pieces that you've done already fit into the form.

And then lastly we'll talk about the risk evaluation process leading to prioritization of next steps.

In the prior video I showed you how you can develop a design tree that looks at, in this case, a wood chair.

It was broken into a set of four systems and within those systems there were a number of So this design tree identifies every piece of your system and then you needed to identify the functions in each piece.

After identifying one or more function for each piece of your design, you then could relate those functions back to the high-level functions, the ones for the wood chair overall.

And that's where we got these three function trees.

After identifying the functions at every level of your design, you could determine failures associated with each one of those functions.

So for example, at the high level, the top level of the wood chair, there were three functions and each function has one or more failures potentially associated with it.

These we call the failure effects because they're the things that the customer experiences.

In addition, each system is going to have its own failures associated with each one of the functions and then each of those systems corresponds or each of those failures for a system are what we call a failure mode.

Failures occur at every level of your design.

So down at the component level, you want to go through each system and look at every component and identify first its functions and then how it might fail to achieve those functions.

So for example, the seat system and the back system, the leg system and the overall general system.

Now each of these red failures becomes what we call the failure causes.

These are at the root of the failure from the design side.

These are the things that you can improve in order to prevent a failure from occurring for your customer.

So now that you have this background info, let's take a look at how it all fits into the FMEA form.

Shown here are the left two thirds of the FMEA form.

We're going to fill this in partially based on the information that I just showed you.

So we'll start out by taking that left hand column and using it to list each system and then each system's function in a separate cell.

So for example, the seat system has two functions.

One is provide seat comfort and the second is to support the user's weight.

And then we move into the back system where its first function is provide back comfort and then we run off the page.

There would be more here if we had more space.

After you've got the system and the functions identified, you then list the failure modes.

Now remember the failure modes are the system failures.

So these are the ones associated with, in this case, in the first row, the seat system.

So the seat is uncomfortable, that's the failure mode associated with providing seat comfort.

Failure mode associated with supporting the user's weight, there are actually two of them.

The seat could break or the seat could flex too much.

That's why the seat support user's weight was broken into two rows.

And then the last one here, providing back comfort, we get back to the back is uncomfortable.

So now we can move on to the third column, which is the failure effects, potential effects.

These are the user experiences associated with the failure mode.

So if the seat is uncomfortable, then the user is uncomfortable.

If the seat breaks, well then we've got a pretty serious issue.

We could have sharp edges exposed, we could have the chair collapse, the user might be injured.

These are some fairly severe potential failure effects.

If the seat flexes too much, then the seat might feel unstable.

And then back to the last row on the screen here.

If the back is uncomfortable, then the user becomes uncomfortable.

So again, we're taking the failure mode is coming from the system level.

The potential effects is coming from the overall design level, the customer experience.

Now we're going to skip past severity because I'll do that in a moment.

Let's talk first about the potential causes of the failure mode.

What we want to do is list the potential causes, which are the component level failures that are associated with the failure mode.

So in our first row, seat is uncomfortable is the failure mode.

The component level failures that might contribute to the seat being uncomfortable are at the seat itself level.

The seat doesn't fit the user, the seat has sharp edges, or the seat has hard spots.

And if you go back in the video, you'll see these are the three failures that we had identified with the seat component.

Moving down to the next row, the seat breaking is the failure mode.

There are a bunch of potential causes for that because there are a bunch of components that contribute to supporting the user's weight.

So the seat might be too thin.

The aprons might be too weak.

The fastener holes might shear.

The corner blocks might be too thin.

The fasteners themselves might shear.

The wood threads could shear.

There might be additional failure causes as well.

So we continue this process to fill out this cause column.

Now we're going to talk about what happens in the severity and further right on this table.

The risk evaluation part of the FMEA process, you're going to look at three things.

How severe are the potential effects of your failure mode for the customers?

How likely is it that the failure cause will actually occur if you follow your current processes for design?

And how likely is it you will detect that failure cause before it gets to the customer, again, with your current processes?

For each of your failure effects, in other words, your overall design failures, rate how important that's going to be to the customer.

And here's a rough guideline that came from statstuff.com.

The 1, 2, 3, 4 are low effect on the performance of the system.

The customer may or may not even notice.

The 9 and 10 are customer safety or regulation effects with ranges in between.

Now considering each of your failure causes, these again are the component level failures that you've identified.

What are your current plans to prevent those?

These would be things like planned analyses, be as specific as you can, and preliminary testing you might be doing to perform.

Once you've listed each of those, then rate how likely it is that the cause will still happen even with your planned analyses and your preliminary testing.

Are you thinking it's very remote or do you think it's fairly high that the cause will occur?

Next, you want to consider how likely it is that you will detect the failure cause before the product gets released.

So what you want to do is list your current plans to detect those failure causes.

These might be customer reviews, if it's an appearance or comfort related item, or it might be planned final testing on a prototype that you're going to use to detect the failure cause.

Then you're going to rate how likely it will be that those methods will detect this cause.

Do you think it's very likely, almost certain, or do you think it's low?

You probably won't be able to detect it before it gets to the customer.

Now let's go back to the FMEA form and we can see how each of those risk assessments fits into the FMEA form.

So the first thing we want to do is go back to that severity column.

Now that you've identified the severity associated with each of the effects, you can list it in this column.

And what we want to list is the highest severity, not the average, not the range, just what the highest one is.

Because you can't control which effect is going to happen when a failure occurs, so you need to consider the worst case.

So we fill each of those in, and then we can move over to the current preventative activities.

Now that's where you want to identify for each of your failure causes, what are you currently planning on doing in your design process to prevent that cause from happening.

So you want to make sure the seat fits the users, so you better measure the users.

See what shape a user's seat has in order to make your seat comfortable for them.

You should also avoid sharp edges, so you round your edges.

You want to avoid hard spots, so you want to smooth out your surfaces.

After you've identified everything you're currently planning on doing, you then want to make that assessment on the occurrence.

How likely is it that this might still occur?

And my confidence isn't incredibly high for seat comfort, so I'm going to give it a 5 for occurrence.

You continue that process down for each of your failure modes.

And once you've got a set of preventative activities and an occurrence, you can then move on to the detection activities.

So how are you going to make sure after you implement these changes in your design that they actually worked?

What is the verification?

Well, for a comfort issue, you might run a customer clinic.

So you might bring a bunch of customers out and ask them what they think.

Do they like it?

Does it feel good?

Those are pretty effective, so I'll give that a 2 for detection.

Moving down to seat braking, undoubtedly you're going to make a prototype and do some testing with that prototype.

That might also be very effective, so you continue to fill in this column.

Once you have detection activities and detection ranks listed, you can then multiply everything out and get an RPM for each of your failure modes.

So all I've got is 1 calculated here because I haven't finished this table yet, but you can see 5 times 5 times 2 gives me 50 for the RPM.

Now the number itself isn't very important.

It's only important relative to the other numbers.

Here's where we get to prioritizing our next steps.

Now that you've recorded RPMs for each of your failure modes, you can go through a prioritization and action setting process.

So to prioritize, first off, you focus on the high RPM numbers.

You're not going to do this for every failure mode, just for the higher numbers, and there is no absolute high or low number we're looking for here as a cutoff.

It is your decision as a team which are the ones that stand out.

Where is your cutoff?

Is it 10% of your total?

Is it 25%?

50%?

The goal here is to prioritize the ones you need to work on, not just to work on everything, because you do have limited time.

For those high RPM ones, you want to identify if there's any way to reduce the severity of the effects.

So affect that S number by making the customer feel it less.

Or can you add some preventative actions to reduce the O number, reduce the likelihood that it will occur for the customer?

Or can you add some detection actions to reduce the D number, so increase the testing that you do?

Regardless of which ones you do, you want to document those actions into the right columns, the recommended actions area of your FMEA, and this is only for the high RPM ones that you're doing this.

Not every failure mode has to have a recommended action.

In that, you're going to state the action, give a responsible party and a target completion date, and then I recommend in another color, identify what you anticipate the change to the RPM to be.

So for example, shown here, let's assume you add an early customer clinic, so you evaluate comfort on your structural prototype rather than waiting to your confirmation prototype.

And let's say that that reduces the likelihood of occurrence from a 5 down to a 2.

That's going to drop your RPM from a 50 to a 20.

That is a major drop just by adding an additional round of early testing.

But you're not done yet.

You now want to make sure that everything you've captured in your FMEA, you are actually going to do.

So I recommend going through the current preventative activities and current detection activities that you've got listed in the FMEA and make sure each one of those is identified as a task in your project Gantt chart.

And in addition, for those high RPM ones where you've added recommended actions, document all of those actions in your Gantt chart.

That's a good way to make sure you're not going to forget about it, and it nicely closes the loop on the FMEA.

Now the last piece of closing the loop on the FMEA doesn't happen until you're ready to release your product.

You're going to go back and document the actions being taken.

So that's the actions taken column that is empty here.

You don't fill that one in until you've actually done what you said that you're going to do.

Once you do it, you can document it there, and that closes the loop.

So at the end of the project, we want to make sure that the FMEA is completely filled in for all of the high RPM items.

That is, you've got an action, you've got a target completion date and a person, and you've documented that you completed the action.

We'll talk about this towards the end of your senior project to make sure you go back and finish this off.

But for now, if you've got recommended actions for the high RPM items, you are ready to move on with the rest of your design activities.