Reliability is Not Metrics, It’s Decision Making
MTBF, KPIs, yield, return rate, warranty… bah!
We may use one or more of these when establishing product reliability goals. When tracking performance. When making decisions.
Goals, objectives, specifications, and requirements, are stand-ins for the customer’s experience with the product.
We’re not trying to reduce warranty expenses or shouldn’t be solely focused on just that measure. We need to focus on making decisions that allow our product deliver the expected reliability performance to the customer.
Reliability Goals, Models, Measures
We use a range of different measures or metrics to describe the expected or estimated or measured reliability performance. It is often a helpful exercise providing necessary insight and guidance.
The issue is when the metrics become the goal, rather than a reflection of the goal.
Improving a metric by changing the method of calculation may appear to improve progress toward achieving the desired outcome yet not actually change anything in reality. The actual customer experience doesn’t care about the measures use to describe the experience.
We can and should use our metrics to help us make decisions and take action.
Not to just to make decisions that change the metric.
Here the focus is on decisions we make when considering the design, the process and the performance of the product.
Reliability is Designed into the Product
I often say ‘reliability occurs at the point of decision’. The design process includes many decisions. Members of the design and development team determine:
- Which vendor
- Which component
- Which derating or safety margin to apply,
- Which coatings
- Which attachment methods
- Which… many other details that make up the design
The different in eventual reliability performance for each detailed decision may be very small, insignificant, or enormous. Sometimes the difference is known and often it is not explicitly known.
The metrics we use throughout the process alone have little impact or influence directly. Especially, when the goals, models, and metrics are not part of everyday discussions and decisions. Selecting a component has an impact on the product reliability. Understanding the impact with respect to the desired goal and customer experience may influence the decision.
Simply listing goals, drafting reliability models, and calculating metric values is not sufficient to actually create a reliable product. It is the decisions the create the product that creates the eventual reliability performance.
Reliability Engineering is About Influencing Decisions
Not about setting and tracking metrics. It’s influencing decisions that impact the reliability performance.
In your organization, ask those around you what the reliability goal is for your product. Most will know or be aware of the goal. Some may even know the importance the goal represents to your customers and organization.
Then ask them how they use that knowledge to make decisions that actually impact product reliability.
If they stumble with a response, you have work to do.
If the answer is vague, you have work to do.
If reliability has a goal if it’s important to your customers and organization, then how does that actually influence decisions?
Your role as a reliability professional is not about setting and tracking metrics. It is about influencing decisions.
The hard part is you are not part of making those decisions directly very often. Sure, some decisions will involve a reliability assessment or study or test. Most will not. Poorly made decisions across the many that are made every day will erode or limit the eventual performance.
- Providing information that enables making better reliability influenced decisions is one option. Design for Reliability guidelines, lessons learned summaries, environmental and use profiles details, etc provide a means to influence everyday decisions.
- Providing training on derating, stress/strength calculations, interpreting vendor reliability claims, and the like on a regular basis increase your team’s ability to consider reliability. One to one coaching, informal training sessions, regular webinars or lunch and learn sessions, guest speakers, are some of the methods you may employ.
- Providing your leadership team with talking points or key questions to ask concerning reliability arms them with the ability to foster interest in reliability performance. Often senior management needs to know how reliability impacts customers and specific business goals. Help them make those connections. When reliability is important than help your team make it visible.
Make reliability a part of every decision of every discussion. Make thinking about the impact on reliability performance second nature across your team. Sure the focus may be on functionality, cost or time to market, yet each set of priorities comes with a balance. In particular, shipping a product that fail prematurely on time is of little value to your customers or organization.
Keep the focus on decisions. Your work as a reliability engineer is more than just metrics, it is about influencing decisions across your organization. From the design process to marketing, including operations, supplier management, sales, and customer service, there are decisions impacting how well your product performs in the hands of customers.
Be a part of decision making.
How do you influence decisions that impact reliability? Share what is working for you. Or share where you are having trouble. Together we can get better at influencing decisions in order to improve reliability performance.