Learning from mistakes 7

[jball1]

I am a structural analyst with two years’ experience as an analyst and two years’ experience as a design engineer. Recently I completed an analysis, wrote the report and sent it to the internal customer (design engineer). He thought that there must be a mistake in my FEM representation of a connection, because he thought the stresses were unreasonably high. The high stresses were going to force a design change which, although small, would be frustrating for the external customer. I kept telling him that the way I modeled the connection was consistent with our typical practice. He kept making the same comment. This was irritating to me, as I was pretty confident that I was in the right. Also, the way he worded his comments was pretty condescending. My model and post-processing had been checked, and so my boss took my side, and overruled the design engineer.

For various reasons, this particular report sat for over a month before it was going to get signed off. Since I knew it was going to sit for a little while, and since I had a little down time this week, I decided to go back into the model to see whether making one small change to how the connection is modeled would affect the results. It was not a change that he suggested, but it was an idea that came to me and I figured I’d give it a try. In looking at the model with fresh eyes, I realized that I actually had made a mistake - not in the way I modeled the connection, but in the material property assigned to the connection. It was a simple, dumb mistake. I fixed the material property, and sure enough, the stresses came down to more reasonable levels.

Looking back, I realize that I should have been much more receptive to his insistence that there was something wrong. He is a far more experienced engineer than I am. I was irritated by the way he made his comments, and the sheer volume of comments he made to my report that were not technical, many of which were simply wrong (note - I pride myself on being a decent technical writer, and he was making many comments that were simply grammatically incorrect. I add this note to head off the comments to the effect that I probably am just a poor writer who needs lots of correction).

Anyways, I spend a decent amount of time on these forums, and have learned quite a bit - both technically and about office interpersonal relationships. I have noticed, though, that many of the threads on the eng tips forums follow this pattern:

1. OP explains the predicament they are in.

2. People comment that they wouldn't have made the mistakes the OP made to get into that situation.

3. People give examples of times they were in a similar situation, but they made better choices than the OP, and it all turned out ok.

Certainly not all threads follow this pattern, but many do. There is nothing wrong with people sharing the good decisions they have made. We all can remember times that we made the right decision under pressure, and things turned out well for us. It can be very helpful to learn from each other's successes. However, I think we can all admit that we have made mistakes in our careers, some serious, and those can be as helpful or even more helpful to learn from than our successes.

What are some of the biggest mistakes you have made in your career - either technical mistakes or interpersonal mistakes. What was the outcome of the mistake? What did you learn from the mistake?

Since the first mistake I shared did not have any serious consequences (fixed before the document was finalized), I'll share another mistake that did have consequences.

Before I worked in analysis, I spent a few years working as a design engineer. In this company, the typical practice was to throw you into the deep end and see if you drowned. I actually prefer this approach, because it is a great way to learn things really quickly. The very first project I worked on was with two other engineers – an engineer with one year of experience and an engineer with about 30 years’ experience who was a few months away from retirement. This was a relatively small project, but had very tight deadlines. The deadline for the first drawing happened to coincide with a long vacation that the younger engineer went on. The older engineer, who had pretty much mailed it in at this point, also went on a surprise vacation at this time. So I was left to complete the drawing and get it out to meet the deadline. I really did not know what I was doing. I had little to no understanding of GD&T, and didn’t even know how to put together a bill of materials. However, I told myself, “If my boss trusts this job to me, I must be capable of completing it, and so I’ll just do my best and then hand it in. Any mistakes will be caught when it is checked.” It didn’t get checked. I sent it to him, he briefly looked it over and sent it out (which is a violation of our company policy, but we’re focusing on my mistakes here, not his) in order to make the date. Our customer immediately caught several of the glaring mistakes in their own review of the drawing, which made us (me) look pretty bad. Even worse though, a year later one of the mistakes showed up when they were trying to build the product. This cost quite a bit of time, money and frustration. It was a relatively small project, and so the time/money/frustration was proportional to the project size, but it wasn’t insignificant. It was definitely painful for all involved. The main things I learned:

1. If I am not comfortable with my work, I should NEVER pass it off as finished, assuming that someone else will adequately check it. In fact, I now assume that no one is going to check my work.

2. The later in the game that mistakes come out, the more money, time and heartburn it takes to fix them.

3. Finally, I learned that I should never ever send something out just to make a schedule. Schedules are important, but product quality is much much more important. When you are about to miss a date, everyone is mad at you. If you are weeks, or even a month or two late, the pressure intensifies. However, once you get the product out (at least at my company), all is forgotten with regards to the date. At that point, all that matters is the product quality. So it is much better to take your licks for being late and just work steadily to complete a quality product than to get a hastily finished product out the door on time, briefly look like a hero, and then spend months/years paying for the mistakes.

Alright, now it’s your turn. If possible, I’d especially like to hear from some of the more experienced engineers who I am sure have some significant battle scars. Thanks for reading and in advance for sharing.

 

[GregLocock]

LPS for the snorgster, yes exactly. You had the right concept, but erred in detail under pressure. Hopefully your engineering nouse would have kicked in when you designed a PV 1/10 as thick as anybody else, you'd have self-corrected, and your unblemished record would continue.

I must admit one job I'd have loved would have been at MetalBox, trying to figure out exactly how light we can make soda cans, and still punch them out at a zillion units an hour.

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[Tomfh]

We share our errors, near misses and other learnings formally

I envy you. In my office (and profession (structural) more generally) I find most engineers are really proud of themsevles and don't admit readily to mistakes, especially big mistakes that count

I'd love to have access to all the cases where indemnity insurers have been notified, so i can learn the biggest pitfalls of the profession, but it's all pretty hush hush from what I can see.

I went to one presentation where an engineer opened up about a major failure of an underground carpark. He was frank about the design errors. It was fascinating. I tip my hat to engineers who are big enough to do that.

 

[3DDave]

One high-water mark is William LeMessurier, the structural engineer of the then Citicorp building. At the risk of enormous financial risk and possible career end, he considered what an undergrad student told him. The article has more information:

http://www.slate.com/blogs/the_eye/...that_could_have_wiped_out_the_skyscraper.html

Personally I managed to mess up one thing that was particularly memorable when the majority of the group kept referring to the operating limits rather than the actual limits. The actual limits included some overrun, but after hearing the operating limits repeated about 500 times by others it just swamped the actual limits in my memory. I ended up using the operating limits to calculate clearances. It was a short-term, high pressure event, with most of the project depending on my finishing. By the time I figured it out there was no means to fix it; all that could be done was to hope the default clearance that was supposed to remain at the actual limits was enough when transferred to the operating limits. I went to my boss with it and he tried to make me feel better about it. It's been 20 years and still bothers me.

Another that was minor was a spring-preloaded mechanism that I thought could be checked with a dial indicator and a push-spring-scale. The instructions were simple, dial up the pre-load screw until the spring scale doesn't move the mechanism as indicated by the dial indicator. Sigh. That's when I found that -everything- is a ^$&#*^ spring. I didn't mind the walk of shame when the assemblers contacted me because I knew they were doing exactly as the instructions said. My plan B took the variability of the process from -50% to +200% to about +/-10% and simplified the assembly process to a clear, repeatable measurement.

 

[MotorCity]

We are all human and making mistakes is a part of life, regardless of your profession. One big difference is the consequences of your mistakes which DOES depend on your profession. If you are a baseball player and drop a flyball, yes its a mistake...but who cares, its a game. If you are a doctor and make a mistake, the consequences could be fatal.

One of my favorite quotations is "An error does not become a mistake until you refuse to fix it".

Be open to the constructive criticism of others and realize that they have the same goal as you do - to get it right. They are not out to humiliate you. Learn from your mistake. Each time you make a mistake, try to take something away from that experience and put it in your "toolbox" so you don't repeat the mistake.

 

[IRstuff]

We had a problem with a processor that was too slow, so we were making small tweaks to see if we could goose the speed a bit. We tried a design rule change on one set of transistor and got a successful rule, but it was clear to the product engineer, me, that this was more of a fluke than anything else and that we needed to continue looking for a more robust solution.

Our GM thought he was smarter, so he committed the entire production flow to the new mask. Needless to say, we wound up with $750k of slow processors that didn't meet spec. He eventually got fired, although it wasn't obvious that this was the reason, since he was a bit of d*ck as well.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529

 

[Sparweb]

I make mistakes all the time!
(Just human, I guess.)
My big mistake story fits in the category of "measure twice, cut once":

It starts with a customer that needed a long cylindrical instrument installed inside the wing of his aircraft. It was part of a much more complex project I was also working on, so you could argue that I didn't give this part of the job the attention it deserved. The instrument was long enough it wouldn't pass through the access panels that were already in the lower surfaces of the wings. The aircraft is a modern carbon-composite design, so putting an access panel into the wing's underside was tricky, needless to say. I started the job at the aircraft, examining the wing and taking measurements. Later in the project, I went to the equipment operator's facility in another city, where I measured their instrument. I went back to my office (on the other side of the country) and proceeded to design the access cut-out. I had all the drawings and structural analysis done before the work began on the cut-out, plus I had negotiated with Transport Canada the conditions of a test plan that would demonstrate that inspection could catch any serious forms of damage. The customer finished the work on the wing, I flew out to do my tests (plus others related to the rest of the equipment on the project) and at last the plane was good to go. It was flown to the city with the equipment operator, where finally the equipment including this instrument was installed. Except... they couldn't. The instrument didn't fit in the hole. The hole was 1 inch too short.

I had made one mistake transcribing my notes into the dimensions of CAD models that I then used to "test fit" the instrument on the drawing. Never realized it until I got that phone call. The facility that did the work on the aircraft wing wasn't able to test-fit the actual instrument either, since it stayed in the possession of the equipment operator in another city. For a long time during that project, I was the only person who had seen all 3 things: the aircraft wing, the instrument, and the intended cut-out to install it, so during all that time I was the only person in a position to double-check all of the dimensions; which I obviously didn't do. I had to re-do the stress analysis, but it was the rework of the wing that really cost the $$$.

Ironically, my family motto is: "I mak siccar"

STF

 

[Belgiancadengineer]

I think it's QC 101 that you never rely on 'one guy' to have to check everything. "If you didn't want your wing re-worked, then don't expect me to never make mistakes" would be my approximate exclamation.

I have been in a similar position with a sheet metal part, on the flat pattern it was missing half the cut-outs, well that's just too bad

 

[Terratek]

The higher the position, the bigger the (my?) mistakes. I'm just gonna leave it at that, lol.

 

[Belgiancadengineer]

That's why hierarchies are a bad way to distribute governance. Governance should be everyone's responsibility and it should be assured through clean, and efficient processes. In most companies it's 'ensured' by paying some higher up manager a fat sum of money. That works too, most of the time but when it fails he gets sacked and some other poor basterd has to take his place. At least he'll get paid well until he makes the inevitable mistake.