Reliability Test Planning
Before jumping in at the deep end it pays to take an holistic view of your product life-cycle. What exactly are your goals for this product, architecture or system? Reliability Test Planning, and reliability planning in general are conducted to formalise design, production, and field reliability requirements to ensure everyone on your team understands the driving force behind your reliability development.
Research & Development funding
Did you know that if you’re based in New Zealand and have demonstrated reasonable success you’re able to apply for funding for development testing through the Ministry of Science and Innovation (MSI)? Funding can be used for development testing only, so it’s perfect for reliability where we’re not certifying the product, we’re improving it.
It’s all about metrics
Without metrics in place how will you know when to stop testing? When to say you’ve won the war and the product is ready for release? Or is in control? The process of designing a reliability test plan will allow you to actively engage key stakeholders during the concept and initial development phases to determine an appropriate level of reliability in the field, target yields for production and key test milestones to achieve prior to release.
So what’s involved?
In order to develop a reliability test plan we’ll work with you to determine some key market requirements. Questions that need answering include:
Where is the product going?
How long is the design expected to last?
How long does your warranty need to be?
What do your competitors achieve?
Are you a market leader, follower, or are you designing on the bleeding edge?
What is your design history like?
Do you currently have field & production data to determine real failure rates of existing models?
Under which foreseeable environments will your product need to operate?
And therefore, which stresses is it likely to encounter?
What is sample availability for testing like?
And so on. The answers to these key questions will essentially form requirements that shall be implemented and tested during product development, manufacturing and field use.
Reliability Test Plan :: Integration
Once we have the basic skeleton of a test plan in place we’ll look at where & when during your development process these tasks should ideally be conducted. Tying requirements into deliverable dates, or stage-gates is essential so that everyone understands their workload and expectations. We have the ability to design a comprehensive reliability test plan that is tailored to your organisations current operating processes. We operate on the following key foundations:
Set your goals
During requirement development we’ll look to set key goals for the product, including:
MTBF (Mean Time Between Failure) target – market based
Target reliability of sub-assemblies
Target reliability of the complete product or system
Who will do what (RACI)
Metrics for:
Highly Accelerated Life Test (HALT)
Accelerated Life Test (ALT)
Reliability Demonstration Test (RDT)
Warranty period prediction
What is an acceptable field failure rate
Yes, we all want zero.
No fault found plan
Predict Reliability :: RBD & MTBF
Based on your initial Bill of Materials (BoM) we can determine sub-assembly and product MTBF using an appropriate standard and environment. We can also determine reliability for each block, and therefore gauge whether we’re likely to hit our design goals based on our component selection.
Reliability Block Diagrams (RBD) and Mean Time Between/To Failure (MTBF/MTTF) are two tools we predominantly use, however there is more to consider. At this stage we also need to consider alternate part analysis, what is likely to go end-of-life during this products life-cycle? Are replacement parts available with the same reliability?
Improve Product Robustness :: Highly Accelerated Life Test
Once we’re happy with our component selections we can design and manufacture prototypes and begin to test. The very first thing we always recommend is to thrash the product early in the design phase. For this we use Highly Accelerated Life Test (HALT) to push the sub-assembly or product to failure, determine dominant failure modes and correct these using sound engineering practice.
In addition to exposing the product to temperature and vibration stress we’ll work with you to determine product specific stresses to apply in combination, or separately to HALT.
Quantify Product Life :: Accelerated Life Test
After we’ve made the product robust we’re looking to quantify the effect of expected stresses on product life using Accelerated Life Testing (ALT). The goal is to determine the acceleration factor of the stress on product life. We do this by exposing multiple samples to varying levels of stress, attaining functional results periodically and plotting these using a defined distribution.
Ultimately this is an incredibly powerful step, it provides critical information for demonstrating life, determining warranty periods, and ensuring your dominant failure mode is understood.
Demonstrate Product Life :: Reliability Demonstration Test
After we’ve determined the effect of a given stress on product life we can begin to demonstrate reliability using closer to real world stress-levels, larger sample sizes and known confidence levels. Reliability Demonstration Testing (RDT) is conducted to statistically demonstrate that an acceptable amount of the population will survive a known length of field use under expected conditions.
Calculate Warranty Period :: Failure Rate
Given the results of your RDT we can determine what length of warranty to offer based on predicted field failure rate. In this way we can equally determine how much stock you are likely to need to replace during this period, how much this will cost, and therefore how this will affect your bottom line.
Qualify Product Performance :: Environmental & Climatic
No doubt you have a bunch of international test standards you put your products through currently. We’re talking about all types of ASTM, ISO, IEC and MIL-STD type tests where the goal is to prove the product will survive in its intended environment – though not statistically. These are tests that either pass or fail.
People often jump straight from MTBF to qualification tests, you’re ultimately setting yourself up to spend a lot on re-testing, finger pointing and false confidence. I’m not sure about you, but for me, I want to make sure that by the time I get to these expensive tests i’m very, very confident my product will pass.
Tests such as temperature, humidity, sand & dust, salt fog, altitude and vibration are all part and parcel of qualifying your product for it’s end use. Not only that, from a market perspective quite often these tests are the only ones released to the general public.
Protect the Product :: Package design & testing
It’s all well and good designing a whizz bang robust product that meets all your internal and external product standards. Another aspect entirely is ensuring the package you design for your product provides the most protection at a target cost. Package testing should never be ignored, it’s an essential part of your product aesthetic – and is most likely the very first thing your customer will see.
Resonant frequencies can play havoc with products, as can poor manufacturing of cartons and low quality adhesives. Additionally if you’re shipping any kind of medical, biological or chemical product you have a duty to ensure those handling it along the transport route will be safe.
We believe in measuring the transport network you operate in to quantify what exactly your packaging will need to survive. From this measurement service we can determine an appropriate level of testing to provide confidence the package will do it’s job.
Control Reliability :: Production
You’ve designed a product that is reliable, you’ve improved, quantified, demonstrated and qualified – and now we’re into production. How are you going to control process drift and supplier quality? The answer for most companies is steady state extended burn-in, a time consuming paradoxical waste of time in todays age.
Components are stronger than they were when burn-in was first introduced, even bad components are stronger than traditional in-process burn in stress levels.
Highly Accelerated Stress Screening (HASS), or Highly Accelerated Stress Audit (HASA) are two processes that can help you control production reliability. Either through acceptance sampling plans (HASA), or via 100% screening for high-risk items (HASS).
HASS is tailor-made to the product being tested, we know what a good product can survive, so we determine a screen that ensures we catch a high percentage of weaker units. This process will expose issues in manufacturing quickly, accurately and repeatably allowing you to control batch related failures, process drift and changes in technology.
Manage Change :: Alternate Parts & Design Changes
When a component or sub-assembly goes end-of-life we often need a quick means to verify the replacement (alternate) part provides the same form, fit, function, and reliability as the one it is replacing. Having conducted the testing mentioned in previous sections you have a perfect picture of how the product performs, therefore, inherently you have the ideal benchmark.
Testing is then a matter of iterating through a selection of these tests, with metrics already set during initial product development. Instead of guessing based on form, fit and function alone this gives you the ability to see into the future.
Life Data Analysis :: Close the Loop
Analysing field data provides the perfect means to closing the loop on product reliability throughout the life-cycle. We use this data to confirm our predictions and to provide targeted improvement opportunities and metric adjustments for the products that are still to come.
There is no data like real data, it is incredibly rich and detailed, though oft-times quite hard to get your hands on. We would say that the information you get from this data is well worth the expense of setting up an accurate reporting tool for your customers and agents to use.