Highlights
-
2024-01-17 12:22 So I’ve definitely seen improvement in the information that’s available out in the world. I can’t say I’ve seen a whole lot of improvement in team members. So my personal opinion is that the design team should go through formal FMEA training and of course define what you want it for. Are you going to start with the product and product failures and stick to that and not try to combine it with the things which maybe should be over in a hazard analysis or if you want to do it in the FMEA? I have worked on companies where would do a user FMEA, which has kind of got that same aspect of a hazard analysis, where now it’s more about the user interacting with the device and not with direct failure modes in the device
-
2024-01-17 12:22 Right. The good planning will tell us what the objectives are, what is a boundary and how do we structure it, because we seem to think that one FMEA for design, one FMEA for manufacturing, maybe one FMEA for software can help us do it all. And sometimes it may not be possible because we work with complicated systems, right?
-
2024-01-17 12:24 Maybe what is the difference between the traditional FMA, I mean, non medical device FMEA and the medical device FMA? Because, you see, non medical device FMA, we have columns of severity, occurrence and detection, whereas in medical device, we have risks defined as a product of severity. And the p one and p two. Got you. Wonderful question. I think this is a great question. Thank you for bringing it up. Roger. You want to take that on? Okay, so when I think of detection right away, I’m going to say the two ways I’ve seen of dealing with it. Does this mean, will the potential problem be detected before the design is complete and released to the marketplace, or is this the kind of kind of detection where you have a device which can interact with the patient? And by detection, you mean if this goes wrong with the product, the patient will be notified by it, because the device itself will self detect and warn about that. So that’s a very confusing factor there, which was why I think the train of thought now is to, when we’re talking about product design, take detection out of it. So we’re only talking about occurrence and severity. Got you. And not deal with that or deal with it over in how we’re going to fix the issue. And Jeanette, you mentioned that. I love the way you put it, by the way. I so love it. Medical FMEA and non medical device FMEA. Right. And this statement actually represents what is happening in the industry right now. FMEA is FMEA. Right. But we are using it for the purpose of risk analysis in the medical device industry. Putting p one and p two out there. I think that was your original question. And what I would say to that is very, very interesting topic. And we can go lot deeper into this. I would just say use an FMEA for the purpose of FMEA, which is improving the reliability of your product, addressing potential failure modes, and not necessarily addressing hazards, hazardous situations and harms. That is what you do in a hazard analysis.
-
2024-01-17 12:23 I’m interested, Roger. You spoke about hazard analysis and FMEA. I’m very keen on using the right technique for the right problem. I don’t work in medical devices. That’s, I think, the one area of functional safety I’ve not worked in. I’m working in automotive systems at the moment. I’ve experienced with hazard analysis and risk assessments and FMEA. And for me, they’re different tools for slightly different purposes, ultimately the same aim, of course. And I’m just wondering, what advice can you give on the right technique for the right problem and what techniques you would recommend over the, I suppose, the suite of functional safety activities. Okay. Andy. I’m certainly no expert in this field. As I say, I’ve become to appreciate it more and more in my career. But in my mind, if our focus is product failures, the design itself, the product itself, FMEA is the choice. If we’re talking about situations and events, at least as a first pass around the device, that’s maybe that preliminary hazard analysis. But as the design evolves, you have, of course, different ways that design can fail. But there also may be, to my mind, at least, different hazards that may be combinations of failures, which maybe show up more in a hazard analysis by the team who’s continually out there rethinking about this, that maybe the hazard analysis is a better vehicle for that type of situation. Two failure modes that create a situation that you won’t see over an FMEA. Yeah, and FMEA is a single failure mode, single fault type of an analysis. Maybe use fault tree. Maybe use event tree. And you combine all these techniques to build a linkage between hazards, hazardous situations, and harm. So I think the message here, guys, is let’s not use just one tool for every job out there. Find the right tool for the right job, and that takes a little bit of work and experience.