In estimating proof test coverage we first need to understand which failure modes can never be detected and think about why we have designed a system with undetectable failures. We should be designing our systems to enable full testing. We also need to think about whether those undetectable failures can be expected to occur at a fixed and constant rate. Many undetectable failures are systematic so are characterised by a probability and not a failure rate. For instance we can't detect that a fuse is the wrong size because we can't test it, but its probability of failure is not likely to increase with time.
In estimating proof test coverage we first need to understand which failure modes can never be detected and think about why we have designed a system with undetectable failures. We should be designing our systems to enable full testing. We also need to think about whether those undetectable failures can be expected to occur at a fixed and constant rate. Many undetectable failures are systematic so are characterised by a probability and not a failure rate. For instance we can't detect that a fuse is the wrong size because we can't test it, but its probability of failure is not likely to increase with time.
Thank you Sir for explaining in quantitatively to understand the basics easily.
Succinct and very well presented.
Excellent sir, thanks for your efgort
Thank you, sir. :)
Thank you