The fist part of this post you likely already know. It’s the second part that may be helpful.
I love HALT testing and almost always include it in a new program. With a team new to the concept there is always the hurdle of getting them to understand it’s value. It’s not intuitive to see value in destroying a product with stepped stresses. Often these stresses aren’t even apart of the product’s use case. Why vibrate a lab electronic device that spends its entire life on a bench? Seeing the failure mode is a capacitor flying off the PCB at 50 G’s doesn’t reinforce the value of the activity without some explanation.
The value in finding out that Capacitor C5 is the first to break loose under physical duress is that it very likely will be the first component to break it’s electrical connection if the board is constructed under a bad solder flow. It also may be the first component to break when the device gets dropped three feet onto a tile floor only once in it’s 10 year life. By finding this first high stress failure early in the program we can make simple changes to up robustness. The solution could be as simple as putting a drop of glue on the capacitor to tie it to a neighboring component. That is a big difference in strength with, what is likely, a free design improvement.
As I said I know many of you reading this already understand this. So why am I writing this post?
To share an education step that may help you explain the value of HALT to design teams.
I always start any HALT program by saying that “The acronym of HALT is wrong, and even Dr. Greg Hobbs (the creator of HALT) said so”
HALT stand for “Highly Accelerated Life Test.” I have a few problems with this. It doesn’t’ provide much information about life. Vibration based failure isn’t much help in predicting life of that bench bound device. You also could debate that it is not really a “test.” Tests are things we want to pass or use to make quantitative statement. If you “pass” a HALT test you have actually failed. What did you learn if there are no failures to root cause? When this happens I often advise we start over with a different set of stresses.
Is it more fitting to replace “Life” with “Failure” and “Test” with “Exploration” Highly Accelerated Failure Exploration possibly.
Let’s do a HAFE! (doesn’t have the same ring to it)
Next time you are educating a group on the value of HALT start by breaking down the acronym, then build it back up based on what you discussed. Who knows what you will come up with, but surely it will be better.
-Adam
John Bessman says
How about the GANDALF test: Generate Additional New Designs by Accelerating Likely Failures. And it’s easy to remember that “YOU SHALL NOT PASS!” the GANDALF test. 🙂
Adam Bahret says
OMG! That is hilarious. PERFECT!