Are engineers insane?

11/3/2011 2:18 AM EDT

The rebut to the quote origin (Rita Mae Brown?) can be doing the same thing over and over again, and expecting the SAME result EACH TIME is impossible!

11/3/2011 6:15 PM EDT

In response to the question "Are engineers insane?", my answer would have to be "Only the crazy ones ... and maybe a couple of the nutty ones too"

11/3/2011 6:47 PM EDT

No, not insane at all. I agree with Bill that this maxim is simply not true. Whoever came up with it obviously had no experience with experimentation, the scientific method or probability and statistics. For that reason, I seriously doubt this maxim is attributable to Einstein.

11/3/2011 7:00 PM EDT

A good engineer will follow the scientific method and do the same thing over and over again, changing only ONE thing each time.....

But as I often tell people who come to work at my employer...
"You know that saying, 'You don't have to be mad to work here, but it helps'?"
"Yes....?"
"Well you DO have to be mad to work here...."

11/3/2011 8:38 PM EDT

You have to remember. Sanity is just social concensus. You are only considered insane when you deviate from the socially accepted "normal".

Since engineers "are" different from most other people, it is no doubt that the non-engineers consider us insane.

Personally, I would rather be insane and have the ability to build things than to be sane and dependent upon the insane to build the things I need.

Think about it.

11/3/2011 11:58 PM EDT

I agree, that "maxim" is not true. Scientific method involves repetition of an experiment to see if it turns out differently next time, or if the original results are confirmed. Sometimes two different groups get opposite results (depending on who is doing the funding, it seems).

11/4/2011 12:33 PM EDT

You wrote, "... are guided by wave fun and probabilities". Boy, what a Freudian slip. Wave fun. Ah, yes indeed.
They say, (whomever 'they' are), that only the insane are truly at peace, or, my translation, can truly have fun. If that's the definition, then I am absolutely certifiable since engineering is just plain fun. I enjoy learning, inventing, and just plain product development.
Now about sitting in the lab waiting for that rare event, may I quote Douglas Adam's Hitchhiker's Guide to the Galaxy? "You're never alone with a rubber duck." That's a good time to let your mind wander (at least, partially) while you work on either the next big thing or something so totally ridiculous that it just might become the next big thing.

11/4/2011 12:36 PM EDT

Generally speaking, scientific result is an universal true. Engineering is built on science. A carefully engineered device has to behave same way every single time. To accomplish the goal of building a product, an engineering organization will spend huge amount of time to repeatedly engineer based on the same technology and QA with thousand of test cases. Repetitive work may drive somebody crazy; yet, it is still far from insanity. Seriously, the success of any business is to repetitively doing similar thing, if not the same thing, over and over again. It is overly good and better than anyone doing it.

11/4/2011 12:40 PM EDT

Using "fun" instead of "function" is just my small way to see if people ar reading this column carefully. Plus, maybe those atomic particles are having their fun by driving us crazy, we don't really know!

11/4/2011 12:50 PM EDT

You've all made a good case against the so-called insanity definition. Another silly platitude is "A good manager always hires people smarter than himself." There was an excellent Dilbert about that one. My take on it is that the person who came up with it should have said, "A good manager isn't afraid to hire people smarter than himself", otherwise there's an ego problem (but that never happens to any of us, right?!). That "always", as Dilbert pointed out, would lead to an organization with the dumbest person at the top!

11/4/2011 12:54 PM EDT

"... those atomic particles are having their fun by driving us crazy, ..." Bill, that's just CHARMing and QUARKy at the same time.

11/4/2011 1:32 PM EDT

IMO, the "expecting different results" bit is the key here. Of course Scientists and Engineers repeat experiments to validate or invalidate hypotheses, and that stands at the core of science (repeatability is key). But some people keep doing exactly the same thing over and over and EXPECT a different outcome every single time even when they are faced with the same results. I have seen this in real life, and I think it is stupid. It is more wishful thinking and superstition then sound scientific thinking.

11/4/2011 2:15 PM EDT

And there I was all these years thinking I was insane!

11/4/2011 3:01 PM EDT

On a micro level, experimentation and debugging are all about small, incremental changes. Changes so small, in fact, that an outside observer, an in observant observer if you will, might not notice any change.

Therefore, there's a method to our madness.

11/4/2011 4:10 PM EDT

blah blah.... what you are talking about is not doing the *same* thing over and over..... but doing almost the same thing, so I don't see how that disproves anything about the truly insane or the validity of that maxim. Slow day at EEtimes, huh?

11/5/2011 2:12 PM EDT

I agree with the maxim. Doing EXACTLY the same thing over and expecting different results is crazy. You want the same results such as in a manufacturing flow. Same thing when proving an experiment is repeatable. When something goes wrong and your quality goes down, you look at the process to see what has changed, not to see what you are doing the same.

11/5/2011 5:33 PM EDT

I'm with those who say it's all about semantics.

The scientific method tries for identical results with the same set of inputs, of course, but repetition is needed to ENSURE that the inputs were the same. You have to convince your peers that you in fact controlled all of the variables, and/or that uncontrollable variables have no effect on the result.

Hardly insanity.

On the other hand, I'm not trying to be overly defensive about our sanity ...

11/5/2011 7:43 PM EDT

If you have ever attempted to start a balky engine with a rope-type recoil starter, it looks insane to those who expect the engine to start first pull every time. IT is far from insanity, it is persistence. Besides that, random variability in things that we can't see is always changing things. So it is not the same thing every time that we do it. Of course, there are some dumb choices that don't work, and will never work, and to try those repeatedly would be insane, but not from expecting them to work, but from trying dumb things repeatedly.

11/5/2011 9:14 PM EDT

Just because it is a cute saying attributed to a bright person does not make it true. Even if Albert Einstein did make the comment it has been taken out of context. It might have just been a throw-away flippant comment.

Executing the same procedure repeatedly is often a very valid thing to do in testing. Except in software in a highly constrained environment the condition, and thus the variables, are often changing. Even as trivial change as time can be a factor.

Changes can be very subtle things that we don't notice. Therefore we can often change variables we don't even know about.

I know from extensive file system testing that repeatedly running the same test for hundreds of thousands of iterations can be valuable. I've had tests run flawlessly for over 50,000 cycles before failing.

Similarly I've run some simple cyclic tests on the Linux kernel which have run for three hours before failing. Having worked on some commercial OSs I know the same problems can happen on those too.

Sure if you have multiple different tests available then run others. But if you've exhausted those tests then just run more of the same.

One of the benefits of automated continuous integration testing is that blindly runs the same tests over and over agin. It does not make biased assumptions that we do and skip over some tests because we obviously didn't make any changes impacting that part of the system.... and yet did manage to change something.

11/6/2011 1:35 AM EST

I agree with you.
In my earlier job, we had developed a TCP/IP gateway for a fault-tolerant
computer used in trading. The architecture was complex, involving a multiple processors, and lots of code.

We had a "reported" problem, that took down the gateway at the customer. The problem was not easily reproduced. I wrote lots of scripts, and wrote a master controlling script to fire off a huge number of concurrent sessions, letting us see the error in 2-3 days, verse about once a month.

But when it broke, no where could we see a error, it just seem to stop working, and no error code at any of the many levels. Step and repeat.

After hammering this multiple times, I finally detected a minute pattern. Everyone thought I was insane, because the pattern I detected was a thermal variation. Ie the room was slightly warmer when it tended to break. (In a data center temp is tightly controlled). Most people think of
me as a software person, but the reality is my hobby is rather serous electrical engineering.

So I brought in standard thermal debug tools, my hair dryer, and can of
cold spray/thermal spray to cool things off.

You should have seen the looks on the faces of the engineers, is he crazy,
is he insane (The cost of one gateway was well over 20K US$). And proceed to power up the hairdryer, and startup the scripts, and blow hot air over the board. Sure enough could break it on command. Even better, I could fix it with cold spray.

Isolated it down to a bicmos dual port static ram, evidently the design engineer had forgotten the hotter the part is the slower it gets. And under
heavy load it would miss a command between the components. Added a heat-sink, which prevent the problem from happening, and change the BOM to have a speed grade faster. Design engineer still says the original was ok, but reality wins in this case.

11/7/2011 9:21 PM EST

I totally agree with you re: good engineers. One software engineer that used to work here simply said to push the reset button when I had pointed out that his system locked up on me. Now that is insanity.

My wife never uses the word insane to describe me but frequently refers to me as weird, different, or strange. She is very kind and I appreciate her for that. 45 years of marriage to a guy who can fix things has encouraged acceptance.

Let's face it, as others have said in these comments, we engineers ARE different. (I've got a double whammy on me, being an analog engineer.) I present one simple example. I go into a room and turn on the light switch, hear the audible click, and then the incandescent light comes on after a 0.5 second delay, and I stand there and stare at the switch. I say there is something wrong with that switch but all she and others can see is that the light is on and I am being unreasonably suspicious and a little on the crazy side. I cycle the switch, it works perfectly, and now I'm really concerned. Everybody "knows" I'm crazy now. Anybody else been there?

I can "see" inside that switch and can attribute the initial 0.5 second delay to a corroded contact, gummy pivot, melted rocker support, weak spring, etc. In my mind the system is broken even though the light is on.

I have told others innumerable times that intermittents are the bane of my existence. An intermittently functioning system simply means that there is something I don't know yet, an undiscovered variable, and much more testing to be done.

My analog circuits are always a part of a digital system. 20% of my work is actual circuit design and the other 80% is figuring out how to test the system in the most realistic way, repeating the tests until it fails, debugging the failure, repeating the tests until I have to finally decide when I'm reasonably certain all of the bugs have been worked out. BTW, the system testing must ALWAYS include temperature cycling.

Marv Amerine

11/8/2011 12:38 PM EST

Since most scientists and engineers exhibit traits that are called either "autistic" or "attention span deficient" by many social "scientists" (and many leading Republican presidential candidates), we need to look at the root of the word "research". As we all know, it comes from the common experience of observing a phenomenon, then--when we look again, it does not happen. So we settle in for a series of increasingly controlled tests in search of the effect and its details. Or in the common parlance, re-search. Michael Current

11/8/2011 3:04 PM EST

Is there any other group of humans that routinely discusses its 'otherness' in this way? Did this weird group escape persecution by its the ability to be useful? Any other group would have been eliminated centuries ago.

Without the kinship of other weirdos geeks and general good-with-their-hands folks many of us would feel estranged, alienated or simply lack confidence through not 'fitting in'. Thanks, EET, for this platform...

11/8/2011 3:34 PM EST

This article is so wrong. If you repeat the exact same thing over again with every atom in the exact same place and nothing changes whatsoever then no, you will get the same results. It's like watching a video on say a VHS. If you watch it over and over again then it will play the exact same video but after a while the quality will fade b/c of wear and it will technically be playing something slightly different. BUT that's b/c there was a change. The tape got worn, meaning that you didn't do the exact same thing over.

What this article is talking about is just changing or finding tiny variables to fix as you repeat an experiment.

Repeating an experiment the same way twice is technically impossible so that should be the topic of this article. "Is repeating an experiment, what seems to be the same way, worthwhile"

11/8/2011 5:17 PM EST

Sorry, Dr Hawkins, you know as well as anyone that atoms and their constituents don't have the same exact place next time around. They can't, nor can we know where they actually are. I guess it depends on what you mean by "exact same".

Think tunnel diode, for example--depends on probablistic quantum tunneling.

11/8/2011 6:27 PM EST

KB3001 got it right. "Expecting" is the key word.

In the example, looking for ppm defects requires a lot of testing. You are expecting the same result, that is a PPM failure, so you have to run a lot of tests to see it.... Any time you look at a distribution you need more than one point to define the result or a shift in the result.

11/9/2011 1:47 AM EST

Engineers are definitely not insane but yes they are passionate. If they are after a problem untill it get fixed they seem lost.

11/9/2011 6:16 AM EST

I think what Bill is talking about is a special case of insanity that makes us not insane. But the insanity as seen by Einstein in my mind is that appears due to the way of thinking of an experiment or an expected results not the way of performing that experiments.
May be he said that because he didn't believe in quantum theory, but in general my answer could be Yes about 90% :)

11/9/2011 2:43 PM EST

You don't have to be insane to work in the electronics industry -- most employers will train you!

11/9/2011 5:15 PM EST

In the land of the insane, only the sane are crazy.

"The definition of insanity is the firm belief in a consequence that is counter to experience." from Sparky_Watt - I think it's pretty common for breakthroughs to come exactly from trying to achieve a consequence counter to experience or common logic. I can't count the number of times I've heard the expression "It won't work. We already tried that." As often as not, the person making the statement was proven wrong.

11/9/2011 10:25 PM EST

In our land enigneers are not considered insane on the contrary they are held in high esteem. The value of education is understood and doing the same thing over and over gain may sound like crazy but thats what work many a times.

This adage falls flat in sports, doing the same thing over and over is what would have contributed to the result getting different over a period of time. The result goes from being a No-man to a Champion.

11/10/2011 6:18 AM EST

From Wkipedia-- Miss-attributed also to Ben Franklin:
The definition of insanity is doing the same thing over and over and expecting different results.

Misattributed to various people, including Albert Einstein and Mark Twain. The earliest known occurrence, and probable origin, is from a 1981 text from Narcotics Anonymous.

Martin Bauer

11/10/2011 6:20 AM EST

According to a book I had at hand, "The Ultimate Quotable Einstein" edited by Alice Calaprice, the quote is indeed misattributed to Einstein and is indeed from Rita Mae Brown, in her book "Sudden Death" (New York, Bantam, 1983, pg. 68).

Brad Wood

11/10/2011 1:58 PM EST

I am going to agree with David Ashton and rogerrobie68. While to the casual observer and one in the same time reference it may look like we are doing the same thing over and over, though usually some change has been made to the experiment. Einstein did not say that. Engineers are often dreamers and dreamers are often considered nuts especially when the dreams are very weird. I thought the article was interesting.

11/10/2011 7:51 PM EST

It does seem that sanity may just mean fitting in with the others. Unfortunately the same different thinking and point of view that make me a good and creative engineer also mean that I very seldom think in the same patterns as all of those who march in lockstep with the lead sheep. If everybody held the same plain vanilla way of thinking then there would never be any progress, or any change at all, except for accidents. So of course, some of us engineers must be a bit insane in order for anything to improve. Probably that means that we must also be rather discontented as well, since if we did not feel any reason to change things, we would probably not bother. "why waste resources?", after all.

11/16/2011 8:56 AM EST

"Doing the same thing over and over again, expecting different results" may be one person's definition of insanity, but it is another person's definition of collecting statistical samples :-)

12/9/2011 11:51 PM EST

Doing something again and again is whats called practice. And practice makes perfect!