In a single meeting, you may need to structure a reliability model, create estimates, outline test plans, and discuss a field failure. The breadth of tools and knowledge to be effective is staggering.
No two problems, questions, situations, or industries are the same. Thus, the solutions you provide must differ as well. If you enjoy a complete set of reliability engineering tools at your disposal, you are well situated to address any question.
This article briefly introduces a few different tools for different types of problems. Consider this my recommended starter tool box.
Reliability program and project planning
Planning is not a unique tool for reliability engineers. Like other uses of program or project planning, you use planning tools to organize the available resources to identify and solve reliability related problems.
A reliability program is a supporting structure for the specific reliability activities necessary for oversight, monitoring, professional development, and project specific support.
Reliability assessment, training, and data collection and reporting are examples of program level tools.
A reliability project focuses on one product under development, manufacture or purchase. A reliability project often focuses on providing reliability information to the rest of the development or procurement team. Reliability models, estimates, FMEA, and accelerated life testing are examples of project-level tools.
There is overlap and your organization may or may not have clear boundaries between what set of planning supports the organization in general, from those that support a specific project.
Reliability risk analysis
Engineers are problem solvers. The first step is the need to be aware of the problem.
Reliability problems with a design or installation may occur quickly or take an extended time to occur. The ability to identify the likely failures provides the entire team the insights they need to evaluate and improve the product or process.
FMEA and HALT are the classic risk analysis tools. Some other risk analysis tools include:
• Simulations
• Finite element analysis
• Material characterization
• Literature searches
• Prototype failure analysis
• Beta testing
Don’t forget engineering judgment. If you or anyone on your team has a ‘gut-feel’ that something is a reliability risk, it’s good practice to investigate the hunch.
Another common approach is to identify anything that has changed, such as:
Material
Vendor
Assembly process
Assembly location
Customer use profile
Component size
Changes from an existing product often include a rich set of new risks.
Design for Reliability
This is a rich set of tool and your design and development engineering teams are the ones to employ them. You may have to provide training and support, yet most of these tools are aimed at making good design decisions concerning reliability.
Since the decisions occur across the design team, and you as the reliability professional do not make most of these decisions, these tools are best employed by others, not you.
Mechanical and electrical engineers should use stress/strength or derating analysis, respectively, to design in an appropriate amount of robustness.
The team should have access to comprehensive information on environmental and use conditions. It may be necessary to create sets of conditions for the different classes of use the product may experience.
This information allows the team to select materials, components, and design solutions that will meet the various environmental or use sets of conditions.
Reliability modeling is one you may need to use, yet encourage your entire team to engage with the model regularly. Modeling may require a simple reliability block diagram, fault tree analysis, or a sophisticated Petri Net or Markov Model.
Tolerance analysis may include worst case, root mean soured (RMS), or Monte Carlo simulations to accomplish.
Statistics to understand the variability within and between:
Environmental conditions
Use profiles
Materials
Suppliers
Assembly processes
Storage and transportation conditions
Installation processes
Statistical tools include the understanding and use of Statistical Process Control and Design of Experiments.
Estimating Future Reliability Performance
This set of tools ranges from an educated guess (engineering judgment) to detailed physics of failure modeling. The aim is to use the available information to create an estimate of a product’s reliability performance when placed into service.
The common tools include:
Engineering judgement
Parts Count Predictions (not recommended for use making future reliability performance estimates)
Vendor data for major components
Field data from similar products
Life Testing
Alpha and Beta testing
A common approach is to break down the system using a reliability block diagram or similar tool, then focus on creating reliability estimates for each subsystem using the best available data for that subsystem.
Addressing Failures
Failure happens. When a failure does occur use the best available failure analysis tools you have available to garner as much information about the failure and its cause as possible.
The 8D or Eight Disciplines failure analysis process guides you and your team from failure detection to celebration of the effective resolution. Root cause analysis is just one part of a failure analysis.
A common tool is some way to track from detection to resolution each and every problem that arises concerning product reliability.
A failure reporting and corrective action system (FRACAS) may be as simple as a whiteboard or as complex as a relational database.
Having Fun as a Reliability Engineer
This is just the starting tool kit. The tools you use on a regular basis will likely be different than my list.
If any of these tools are new to you, then you have something to learn. If you’re not sure how to use a tool, then look for an opportunity to give it a try. If all of these tools are familiar to you, then what have I missed in this list.
What would you recommend be added or removed from this starting reliability engineering toolkit?
Richard Robertson says
Actually, you need to be aware of risk and problem.
Fred Schenkelberg says
thanks for the comment, nice addition. cheers, Fred