There are two basic philosophies when creating a reliability plan for a new product or system.

One is to experiment with prototypes as quickly and often as possible, the build, test, fix, approach. Or, you can research and model detailed aspects of the materials and structures to characterize the strength of a product or system, the analytical approach.

Both methods have obvious applications and not so obvious limitations.

1. Build, Test, Fix Approach

This idea is to get something in your hands as soon as possible.

This may be a Lego mock-up or a crude electronic bread board. It may only have one feature at the moment and not look like the finished product at all. Or it may be many variations of the industrial design so you can evaluate the look and feel aspects of a design.

With each iteration, you learn a little about what works and doesn’t work. As the process continues the mock-ups or prototypes begin to resemble the final product in appearance and functionality. The learning continues and the design evolves into a final form.

The build, test, fix approach for reliability planning has the benefit of allowing the exploration of failure mechanisms early in the design process. Use risk assessment tools to focus on the areas of highest reliability risk and create prototypes that allow experimentation related to the specific risks. If a potential element of a product is not reliable and identified early, the design may address it early with minimal impact to the design or schedule.

Evolution of the design

The downside of the approach is the changing nature of the design. With each evolution of the design, there may be multiple changes which may improve or degrade the system’s reliability performance. Since many reliability tests take time to accomplish, the results may not apply anymore, or may not provide feedback fast enough to impact the design.

The reliability program must continue to address reliability risk with each change in the product. The team requires creative reliability testing on limited prototypes that provide meaningful results quickly.

By focusing on high-risk areas and testing elements of a product for specific failure mechanisms that address the latest proposed changes, the reliability program can provide guidance to create a very reliable product.

2. Analytical Approach

The other approach for product or system development is to conduct careful design analysis, virtual modeling, and evaluation of material and component failure mechanisms.

This approach relies heavily on simulation tools such as SPICE modeling for electronics, Finite Element Modeling for mechanical elements, and Computational Fluid Dynamics for air or fluid flow.

The challenge for reliability engineering is to have sufficient characterization of failure mechanisms to model the behavior within the simulations tools, or with sufficient knowledge of the use environment. There are plenty of physics of failure models available for select failure mechanisms, yet do they apply well enough in your particular situation.

For novel materials or constructions, you may need to conduct detailed studies to develop an appropriate model. For many simulation tools, you will need to know how material properties change, or how electrical component parameters may drift. It does take more knowledge and does not demand prototypes.

Like the build, test, fix approach, the analytical approach does rely on risk assessment tools to focus the effort. The analytical approach does assume you and your team know about and have sufficient knowledge to fully model all relevant (likely to occur during expect use conditions) failure mechanisms. This is not always true.

This is not always true.

In Practice Use a Balanced Approach

To minimize the downsides of each approach we often use a combined approach to building a reliability program during product development. The advantage is you can minimize expensive testing by using existing knowledge and models.

In practice, a new design builds on existing products and systems. The power supply may use the same technology as well as the display while adding a new module for a specific new feature. The motors and bearing remain the same while using a new material for the housing and structure.

Start with the risk assessment, build the plan using the tools that reduce uncertainty. For novel designs plan to expose new failure mechanisms.

The intent is to provide feedback to the design team with meaningful information. Using testing or simulations are just approaches as you design for reliability.

BIO

I am the reliability expert at FMS Reliability, a reliability engineering and management consulting firm I founded in 2004. I left Hewlett Packard (HP)’s Reliability Team, where I helped create a culture of reliability across the corporation, to assist other organizations.