Software Design and How it Influences Reliability

My first system project was a new TV design by HP.   My first actions were to interview all project engineers and discuss their function in the program.  I wanted to know what they were thinking and the approach they would use.  From this interview process, I could start defining the Functional Architecture (Ref. REI book p 22-25) and start forming the reliability strategy.

As I began interviewing, I came to the Software (SW) engineers and got the following response:  “SW inherently is void of reliability problems (defined as ‘quality over time’); Once it is proven to have high quality, it will stay that way”.   Since I had never dealt with SW reliability, I got the SW engineers to brainstorm what things could change over time.  The following is the outcome of that discussion (these can be found in my book REI pgs 36-37).

How SW quality can change over time:

• Manufacturing level – changes needed for in-house burn-in and diagnostic tests as customer feedback is obtained.
• Customer level –
  • Root cause diagnostic tools for Service/Customer interaction to help solve issues without having to send out service technicians,
  • Techniques to upgrade and fix SW and Firmware (FW) in the customer environment using download tools such as memory sticks,
  • Pro-active warning indicators of degradations and eminent failures,
  • Strategies to allow future applications to be added to the TV without damaging the original SW.
• Data collection tools –
  • Accumulate failure rates and types of failures to plan for spare parts and maintenance needs,
  • Feedback to future products on component weaknesses and environments found in customer applications.

Now that the SW designer understood how their efforts could indeed change the reliability of a product, I obtained their commitments to achieve the desired reliability goals.  This had never been done before at HP and was a great way to proactively build in reliability.

We found out that these efforts were crucial at the introduction phase of the HP TV.  All the areas listed above were encountered several times.  Our strategy to deal with them paid off big time.