TS Newsletter Archive


What is Design for Reliability?

Reliability is one of the most important characteristic a product or service must have, from the customer standpoint. Thus, it is extremely important for the success of any company to ensure reliability on their offering across its entire life, even if it is a service delivered on point, or a longer life asset. So, how can managers make sure the organization will deliver services and products under expected levels of reliability BEFORE they are produced / delivered? An adequate Design for Reliability (DFR) is the answer.

Before going into the process, we must define the basics concepts that will be involved in the DFR effort.

  • Reliability: is expressed as the probability (R) that the item or service provider will perform or deliver the intended function or service. 
  • Confidence: is the probability that the reliability (R) of the product or service provider will be the one specified. 
  • Time frame: is the specified period of time where we define the reliability and confidence related to it. 

Thus, a complete reliability statement would be: 95% of reliability with 85% of confidence at 6 months of operation.

Now, the best way to define the importance of reliability and its benefits to the product or service provider is to identify the cost of not being reliable. It is important to note that most of the costs from being unreliable are not evident at first sight or are not immediate to a product failure or poor service experience; most of those cost are hidden.

The first, evident costs of lack of reliability can be summarized in additional production / service costs; such as warranty, spare parts and repair costs. These costs may be small, but can be significant if lack of reliability remains over time.

The hidden, thus more dangerous costs are not evident or immediate. These can be listed as follows:

  • indirect warranty costs (inventory, over production of spare parts) 
  • problem solving costs (reengineering, root cause investigations, field trips, product recall costs) 
  • opportunity costs (cancelled sales, lower margins on future contracts) 
  • long term business costs (liability/legal costs, lost of potential customers) 
  • and ultimately, damage to reputation and brand image. 

An adequate business strategy must include the Voice of the Customer to set priorities and identify key elements that customers identify as quality drivers, then, internal capabilities should be improved or build if they don’t exist, and finally integrate the entire process into the business culture.

DFR - Five basic steps

1) Set reliability goals.
     a. Identify real customer needs using QFD or similar process.
     b. Establish reliability goals, keeping in mind the complete reliability sentence structure described above.
     c. Generate or adapt reliability metrics.
     d. Define the operation conditions where the reliability will be measured.

2) Describe the system using a useful model.
     a. Identify the individual elements in the system.
     b. Drill down into reliability models for each element, if necessary.
     c. Identify critical components and failure potential for each.
     d. Check for failure accelerator elements (temperature, cycles, loads, rates, etc).
     e. Set a realistic reliability level for each component, using different benchmark strategies.

3) Use model to design and predict responses.
     a. Identify controllable and uncontrollable parameters (Key control parameters, KCP; and Key Noise Parameters, KNP).
     b. Map reliability dependencies and interactions across the model.
     c. Use simulation tools and perform experimental analysis.
     d. Conduct life prediction through reliability modeling.
     e. Develop risk mitigation plans (i.e. FMEA) including visible costs, such as warranty and service costs.

4) Test and evaluate if reliability commitments are being achieved.
     a. Develop and apply exhaustive testing strategies, such as:
         i. Internal (short and long term)
         ii. External (field sites, beta tests)
         iii. Accelerated testing (failure mode test if possible)

5) Implement a control and audit process.
     a. Conduct process verification
     b. Conduct product verification
     c. Establish agile failure detection, reporting and solving system.
     d. Establish a robust data collecting strategy from field and internal processes.

This simplified five step process is a general summary of more detailed DFR processes that can be found in different businesses and management cultures. Each business may also have different strategies to address Design for Reliability for services and products.

In future issues of this newsletter, The Power of Collaboration, we will address some basic concepts related to product Design for Reliability and the key statistical and mathematical models to analyze it, such as the Bathtub Curve, Time Between Failures, Failure rate, Availability, Maintainability, and some basic analytical tools to get the most from these measurements. In the meantime, should you have any question regarding DFR, please get in contact with the consultants at TeraSigma Consulting.

Practical Reliability Engineering, 3rd Ed. Revised, Patrick D.T. O’Connor, John Wiley and Sons, Inc., 1995
Improving Product Reliability: Strategies and Implementation, Levin, Mark, Kalal, Ted T., 2003

No comments:

Post a Comment