We help our customers design circuits & systems that empower world-changing products.
Product failures are bad for everyone. Whether it be the customer who has purchased something that breaks, or the manufacturer or distributor who honours the warranty claim.
So, what makes electronics reliable?
For products, reliability is sometimes measured as mean time to failure (MTTF), and for systems as mean time between failures (MTBF). Unfortunately, the only way to measure MTTF or MTBF accurately is to put a large number of items into service and measure when and how many fail over a long period of time! Since this is almost always unacceptable, and often also impractical, a statistical method is sometimes used to predict failure based on published failure rates of components. A reliable product should have a low failure rate and therefore a high mean time to failure. The statistical method is widely used in some industries, notably aerospace, medical and military. However, there are pitfalls:
- The method applies to random failures between early life and product wear-out (the bottom of the so-called “bath tub curve”). However, the success or failure of a product will often be judged during the product introduction phase
- The relatively low random failure rates found by the statistical method are often dominated by a non-random cause, such as a design flaw or problem in manufacture
At Cambridge Consultants we approach reliability from the standpoint: how can we root out the causes of unreliability which could cause the product to fail?
Putting our approach into practice starts in the design phase and continues all the way through to manufacture. Considerable effort is put in to processes early in the design phase, such as technical specification, examination of possible failure modes, supplier and component selection. Reliability prediction can play a role at this point: although it cannot provide an absolute value for MTTF, it can be used for providing a comparison between different designs.
Design reviews during the design process are used to ensure that any design flaws are uncovered. This akin to what an author does after writing a piece that they review and re-read many times. There is immense value in it being proof read by someone independent, who can spot mistakes that the author has become blind to and challenge assumptions which are not founded in fact. A design review offers the same opportunity to ensure that mistakes are corrected, and assumptions are challenged and justified.
Alongside paper-based review, testing of parts and of the whole product during the design phase is important, but can be expensive. In a later blog I will cover how to construct a verification test process which uncovers causes of unreliability effectively at reasonable cost.
The process of excluding unreliability extends into product manufacture, where asking the question “Where could product unreliability be introduced here?” can find dozens of answers. A physical visit for selection of the manufacturing site and during pre-production or to witness production of a similar product is essential.
Cambridge Consultants has undertaken the role of independent technical review for a number of client-designed products, particularly in industry sectors where faults and failures can be costly to resolve. So, if your team needs an independent technical expert to talk to and get advice from, or if you would like someone to do a design review of a product you are buying or importing, then please get in touch with us.