Engineer mistakes

[hamdi.tn]

Hi,
Well the question is pretty much easy, how to handle engineering mistakes, define responsibility within the team when working in small count teams.
My personnel experience in my work is am the only engineer in place for 3 years, so it's up to me to do schematics, PCB, test, mechanical design for enclosure, EVERY DAMN THING. Project manager is more a commercial than an engineer, so no one to check or validate what i do.
Am not here to  so the point is ,recently  i lost my job for a pcb design error that cost less than 1000$  and am cool about it, am opening my own business and i already have a few client that trusted my work for 2 years now. And as am going to be again in charge of everything , am interested to know from more experience developer their opinion about the right way to handle engineering mistakes, how to manage the team in order to avoid those error effectively in small team environment .

[Psi]

Do things slow enough, with checklists and reviews, so that you don't make many mistakes.
If you're then asked to do something "faster", say 'Sure, i can do, but mistakes are more likely as i will have to reduce some of the reviews and rechecking phases'.

That way it's less your fault and more a known tradeoff due to speed.

And if you do make mistakes you can at least prove you did try to check things by showing your checklists.
Even if the checklist didn't pick something up you can at least say you will add it to the list to prevent future occurrences.

[brabus]

In one word? Review, review, review.

Schedule a review, and DO IT. Make the projects being reviewed by someone else in the team.
After the review, schedule another important thing: second review.

99% of mistakes we found was @ schematics level, with a very few percentage of wrong components' footprints.

Then, finally, learn to live with mistakes. In the prototyping phase, the concept of "mistake" is senseless: the pressure is so high, the timing so tight, that a mistake is basically unavoidable.

I am really glad to hear you leaving a place where they dump people for 1000$ mistakes. I really wish you the best of luck with your new activity! You deserve the best!

[Psi]

yeah, a bad track/component on a PCB that requires a new $500 8layer prototype isnt a mistake.
It's a PCB revision.

Also, make it known to your boss anything environmental that is causing you dev problems like distractions/noisy office etc..
Then when mistakes are made you can call attention back to that as a possible reason for it.

Higher-ups like to have reasons for mistakes.

[AndyC_772]

Mistakes can and will happen, so the first thing to do is accept it.

If you have people in your team with overlapping skills, then it's always worth making sure the work done by one person is reviewed in conjunction with another. Sometimes the simple act of explaining how something works is enough to highlight errors - so it can be just as valuable to have the work of a senior engineer reviewed by a more junior one as it is the other way around. Bonus: the junior engineer gets trained as part of the process.

It also helps that nobody gets, or feels, personally blamed for mistakes. If you design something that has a bug in it, then all eyes are on you. If, however, you design something that gets reviewed by someone else and which STILL has a bug in it, then at least the two of you can share the guilt.

No review will catch everything, though. There will always be instances where a mistake isn't the fault of anyone on your team, but instead is down to something external over which you have no control.

For example, I designed in a chip last year which had several pins marked as LED outputs. What it didn't say in the data sheet, was that during reset, they also act as configuration inputs, and that if they're sampled as logic 0 as the chip exits reset, then it goes into a factory test mode where nothing works properly. No amount of reviewing was going to catch that particular bug, and a PCB respin was - IMHO - inevitable.

Try and have a process which avoids silly, unnecessary mistakes, of course - but make sure your team understands that they WILL happen, and plan accordingly. If you can take a rev A board to production then that's fantastic, but it's not the norm.

[Psi]

If you design something that has a bug in it, then all eyes are on you. If, however, you design something that gets reviewed by someone else and which STILL has a bug in it, then at least the two of you can share the guilt.

+1

Misery loves company

[vortexnl]

Sucks to hear they sacked you! But it sounds like you're better off starting your own business
A while back I designed a small circuit board for my work, and even got it looked at by a more experienced electrical engineer

When the first board was soldered, it didn't work. It was due to the ADC on the board sharing the same I2C address as the RTC on our development board, but the requirements were that this board had to work with our development board, so it cost my boss about 100 euro's in PCB costs. He is cool about it though! 

[tautech]

Hi,
Well the question is pretty much easy, how to handle engineering mistakes, define responsibility within the team when working in small count teams.
My personnel experience in my work is am the only engineer in place for 3 years, so it's up to me to do schematics, PCB, test, mechanical design for enclosure, EVERY DAMN THING. Project manager is more a commercial than an engineer, so no one to check or validate what i do.
Am not here to  so the point is ,recently  i lost my job for a pcb design error that cost less than 1000$  and am cool about it, am opening my own business and i already have a few client that trusted my work for 2 years now. And as am going to be again in charge of everything , am interested to know from more experience developer their opinion about the right way to handle engineering mistakes, how to manage the team in order to avoid those error effectively in small team environment .

Unbelievable, I hope they replace you with the Engineer they deserve. 
Now that you will be working for yourself there will be considerably greater ownership of your work and as a consequence of that a much lower error count.
If your business succeeds you might consider performance bonuses and/or profit sharing to give your employees a vested interest in your company's success.
Good luck in your endeavors.

[Kjelt]

If your first pcb run is 100% ok you are very lucky, normally there are multiple engineering sample runs needed to test also for EMC compliancy etc.
Without colleagues to review you can use this as an (costly) alternative for your boss to deal with before the project starts.

[Psi]

oh, your starting your own business now.
I must have missed that bit.

hehe you are the boss lol.

[Legit-Design]

Also, make it known to your boss anything environmental that is causing you dev problems like distractions/noisy office etc..
Then when mistakes are made you can call attention back to that as a possible reason for it.
Higher-ups like to have reasons for mistakes.

this


non-geek = boss
geek = anyone who does something other than talk all day

It's hard for some people to understand why it's not helping if they come and ask the same damn thing every 5 minutes.

[hamdi.tn]

thanks everyone for your advices and encouragement

In one word? Review, review, review.

a lesson to learn 

Mistakes can and will happen, so the first thing to do is accept it.

tell that to someone who designed 2 board in 10 years, and they were a total fail in market

it cost my boss about 100 euro's in PCB costs. He is cool about it though! 

well 100€ isn't much at all, an overreaction depend on the boss personality and i think it depend mostly on how much is familiar to design process, or he is just the kind of boss that only think about money.

hehe you are the boss lol.

well, yes and when i think how engineers are treated in industry i hate being the boss, and the main reason am asking for advice is to be fair to people who will eventually work with me.

[hamdi.tn]

Also, make it known to your boss anything environmental that is causing you dev problems like distractions/noisy office etc..
Then when mistakes are made you can call attention back to that as a possible reason for it.
Higher-ups like to have reasons for mistakes.

this


non-geek = boss
geek = anyone who does something other than talk all day

It's hard for some people to understand why it's not helping if they come and ask the same damn thing every 5 minutes.

love this one , think i will put it on my office wall 

[Terabyte2007]

Mistakes are inevitable. It's how you learn as an individual and work as a team to correct and improve upon the design on an ongoing basis. We run into design issues all the time and I used to see the blame-game being played out quite a bit at my old company which did nothing but hurt them and in the end they went out of business. People need to learn from their mistakes, accept them and improve upon their designs based on them. Without mistakes, we never learn!

To be fired of something as stupid as a 1k PCB error is bad management. I would guess that the company will have a high turnover rate which will effect their success in the future. Usually when a design error is found in our group, it's flagged and any immediate customer matters or service is given priority. A review meeting is scheduled with engineers, management and software developers if needed to determine a corrective course of action. The mistake is fixed and we move forward learning from the incident. No finger pointing, no blame, just a solution.

[Richard Head]

If you expect the prototype PC board to work perfectly you are setting yourself up for disappointment.
I have long given up on that utopian ideal.
Normally my third board is production ready.

[EEVblog]

Am not here to  so the point is ,recently  i lost my job for a pcb design error that cost less than 1000$ 

Wow. That's not a company you want to work for anyway!

[dannyf]

I would put accountability and performance management up there, for a small team.

Accountability requires clear roles and responsibilities. That's where you need to be careful with reviewers - are you better off with more workers or reviewers? If you want to have reviewers, how do you structure them? How are responsibilities shared between reviewers and workers? ...

Performance management means financial rewards and career opportunities, both positive and negative.

You will find that firing people is hard, both for the manager, as well as for the team. But it can be turned into a positive experience (not through fear), by trimming the fat and making the whole team more productive.

You will also find that engineers are far easier to manage, than non-engineering types (like sales or HR).

[dannyf]

i lost my job for a pcb design error that cost less than 1000$

I suppose that monetary losses are far less important in this case than the context under which the mistake was made.

[EEVblog]

I don't recall who said this, I think it was in a book on I think IBM, but true story.
A new engineer screwed up and caused $100K damage or something like that. The manager, when asked if he was going to fire him said:
 "Fire you? Are you crazy, I just spent a hundred thousands dollars on your education!"

[hamdi.tn]

i lost my job for a pcb design error that cost less than 1000$

I suppose that monetary losses are far less important in this case than the context under which the mistake was made.

not for the boss, who asked that i repay the production cost 5mn after we detect problem

Wow. That's not a company you want to work for anyway!

sure thing

[hamdi.tn]

I don't recall who said this, I think it was in a book on I think IBM, but true story.
A new engineer screwed up and caused $100K damage or something like that. The manager, when asked if he was going to fire him said:
 "Fire you? Are you crazy, I just spent a hundred thousands dollars on your education!"

 well here company suppose that recently graduated engineer can built rocket, and experienced engineer do not belong to human kind and more like unmistakable designing god and most of the time they blame you for not being one.

[dannyf]

but true story.

I often tell people that my value-add is that I have lost hundreds of millions of dollars for other people so they don't have to make the same mistakes.

Very effective sales pitch.

[KJDS]

Also, make it known to your boss anything environmental that is causing you dev problems like distractions/noisy office etc..
Then when mistakes are made you can call attention back to that as a possible reason for it.
Higher-ups like to have reasons for mistakes.

this


non-geek = boss
geek = anyone who does something other than talk all day

It's hard for some people to understand why it's not helping if they come and ask the same damn thing every 5 minutes.

At one place I worked at, I'd work from 10am til about 3:30, then from 8pm til 10pm.

I used to achieve far far more in the evening session.

[jlmoon]

I have found that by grouping designs in a gang of  3's.. in other words.. one Engineer designs a device, two others proof / check design drawings, schematics, pcb layout, suggest revisions and cycle through this process until final viable prototype is functional and meets specifications.  Realizing it cost more, but in the long run less mistakes and a much better product or device. 

[nctnico]

i lost my job for a pcb design error that cost less than 1000$

I suppose that monetary losses are far less important in this case than the context under which the mistake was made.

not for the boss, who asked that i repay the production cost 5mn after we detect problem

My answer would be: that is OK IF I get 50% of the shares of the company. Who owns the company takes the financial risk.