Implementing Failure Mode and Effects Analysis (FMEA) in an organization often seems like a relatively straightforward and simple task, but it often hides some intricacies that have to be accounted for. You need to be particularly wary of how the FMEA process affects the long-term performance of your organization, and what improvements you get as a result of its application. By putting enough thought into the fundamental process, you can easily ensure that FMEA will drive your organization towards permanent improvements in multiple areas.

在組織中推行失效模式和影響分析（FMEA），通常看起來是一個相對簡單的任務，但時常會卻隱藏著一些需要解釋的復雜問題。我們需要特別注意FMEA過程如何影響組織的長期的績效，以及應用FMEA之后所得到改善。通過在基本流程中投入大量的精力，我們能很容易去確保FMEA推動組織在多個領域實現永久性的改善。

You must address every critical issue

必須解決每一個關鍵問題

FMEA is not something that you can just apply selectively to specific issues in the production process. You either do it all the way, or you look for an alternative solution that’s better suited for a modular approach to its implementation. You cannot leave any critical issues unaccounted for when running a FMEA iteration. However, this also assumes that you have a good grasp of what exactly constitutes a critical issue in your particular organization, which is why so many companies fail at getting this point right on a regular basis. You need to work with accurate sources of data. Which actually leads us to the next point…

FMEA，并不能選擇性地應用于生產過程中的特定問題上。要么一直這樣做，要么尋求一個更適合其實施的模塊化方法的替代性方案。在運行新版本FMEA時，不能忽略任何關鍵的問題。然而，這也假設是你很好地理解了在特定組織中什么才是關鍵的問題，這就是為何如此多的公司并未能時常抓住這一點的原因之所在。我們需要使用準確的數據源，事實上，這實際上會引導我們到進入下一個觀點…

You have to work with real-world models

必須使用真實的模型

When reviewing your system structure to determine potential points for improvement, you have to keep in mind that simulated, lab data is far from what you’ll be working with in the real world, and you must be prepared to account for that in the FMEA iteration. If you can obtain data sources that provide you with a more realistic model that matches the actual situation on the market accurately, this would be much better than relying on simulated or predicted data, which may or may not match reality all too well.

在查看體系結構，以確定潛在的改善點時，我們必須記住：模擬的實驗室數據與現實要使用的數據確實相差甚遠，我們必須在準備FMEA新版本的時間就得考慮到這一點。如果獲得能給我們提供與市場實際情況精準匹配的更為真實的模式的數據源的話，那么，這遠比依賴模擬或者預測的數據要好很多，事實上，模擬或者預測的數據可能會與實際情況很好地匹配，也可能與實際情況并不匹配。

FMEA must be integrated deep into the production cycle

FMEA必須整合到生產周期之中

FMEA should not be thought of as something that comes after everything else is put in place. A successful FMEA run requires that you integrate it deeply within your actual production processes, and it has to be tied closely to other core steps for an optimal effect. This may often mean that you need to rebuild your production infrastructure from the ground up (or at least to some extent), but at least it will ensure a sustainable long-term solution. Of course, you’ll have to make adjustments to that system sooner or later no matter how robustly it was designed in the first place.

我們不應該把FMEA看作是在其他一切都實施之后所發生的事情。成功的FMEA運行，需要我們將它深入到實際的生產過程之中，并且其必須與其他的核心步驟緊密結合，以獲得最佳效果。這通常意味著我們需要從頭開始（或者，至少在某種程度上如此）重建生產基礎設施，至少，這會確保可持續的長期方案。當然，無論體系起初被設計的多么強大，我們遲早都得對其進行調整。

You need a strict, rigid procedure for FMEA

FMEA程序需要嚴謹且嚴格

A systematic approach to FMEA guarantees consistent results, which is a critical factor when working with the technique. After all, FMEA was among the pioneering techniques in terms of the rigidity and robustness of its design, and this has to be carried over to your actual implementations of it as well. Come up with a plan and stick to it as best as you can. You’ll sometimes have to deviate from the original solution, but as long as you’re learning something along the way and you gain a better understanding of how your company’s systems work together, you’re still ahead. The point of FMEA is, after all, to drive an organization towards continuous improvement in a managed, systematic way.

FMEA的系統方法確保了結果的一致性，這是使用該方法時的一個關鍵的因素。畢竟，就FMEA設計的嚴格性和堅定性，FMEA是一種具備開創性的方法，這一點也得轉移到實際實施之中。請制定計劃，并盡你所能堅持下去。有時，不得不偏離最初的方案，但只要在這一過程中學到一些東西，并且對公司的體系如何協同工作有更好的理解，我們就仍然會遙遙領先。歸根結底，FMEA的目的是：以管理的、系統的方式，去推動組織的持續改善。

FMEA must produce a workable final result

FMEA必須產生一個可行的最終結果

And this brings us to the final point, but one that’s just as important as the others. When working with FMEA, you have to ensure that its implementation actually leads to concrete, workable results that you can put to use within your organization and see a long-term improvement from them. There’s not much point in learning something during an FMEA run if it doesn’t ultimately lead to changes in the company’s structure. Keep in mind that sometimes those changes will be on a smaller scale than you would expect, but there’s nothing wrong with making a less pronounced alteration to the company’s structure as long as it improves something overall.

我們來到了最后一點，但這和前面所列的其他四點一樣重要。在使用FMEA時，我們得確保對FMEA的實施事實上會產生具體的、可行的結果，我們在組織中可以利用這些結果，并從中看到長期的改善。如果最終并沒有讓公司的結構的發生改變的話，那么在FMEA運行期間所學習一些東西就沒有什么意義。請記住：有時這些變化的規模，確實會比我們預期的要小，但只要公司在整體有所改善，那么，公司在結構上沒有出現明顯的改變其實也并沒有錯。