Pawl Bearing's auto design engineering rant

unsexy like automotive engineering changes

2012-02-07T18:13:00.000-08:00

A good friend of mine started working for a start-up company that plans to design and produce an innovative new produce that I swear I've already seen before. They have 6 months to finish the design and then presto they're starting production. After 2 years of production, they'll all be multimillionaires and retire at 45.

Which leads me to the topic of this week's blog: engineering changes.

No matter how awesome you are and how effectively the team you're in works, failure will pay you a visit: screws will strip, chrome will corrode and bearings will break. Or if nothing dramatic happens then you will rudely discover that your transistor is made in the lowest spot in the whole of Thailand and assembled with silicone imported from Fukushima in Japan.

OK so we might run into a snag or two. So?

Dealing with changes is surprising difficult. In fact it separates the boys from the men. Let's consider the most ubiquitous of failures: A screw boss broke in vibration testing. Easy eh?

Well, here are the steps to correct it:

Propose a change, model the change in CAD, simulate the improvement using FEA, get a quote and timing for the change, tell your customer that you have a problem, deal with daily conference calls to update them on the situation, place an engineering change requisition into the internal system, read email that your requisition was rejected because of insufficient proof that the change will fix the problem, change the colors in your old FEA simulation so the "before" is predominantly red and the "after" is predominantly blue, resubmit, submit another change requisition to your customers change system, consult with your materials planning department so that just enough old parts are made, write up a test plan and submit it to the test plan database, reapply for database access because your account was inactive for 3 weeks, update the drawings for the part and release the CAD to the tool and die maker, update the DFMEA, the assembly drawings, the customer drawings, the materials database...

once change is made:

Make sure all old stock is discarded, have new product built, submit approved test plan requisition to the lab, buy them donuts to bring your test to the top of the pile, open email telling you that your part is failing a test it used to pass, promise more donuts if issue just goes away, get tapped on the back for executing the change ahead of schedule and under budget (yeah right).

Experiments to detect that DFMEAs don't deliver

2011-12-09T16:35:00.001-08:00

Like communism, there are 101 reasons why FMEAs should work and why they should result in a better world. In fact the whole concept of FMEAs is pretty convincing until you enter the real world of mass produced designs churned out by overworked engineers. And please don't blame the management of these companies, they're trying to make a profit. Paying a design engineer US$105,000 to spend their time sitting in endless meetings deciding if a jammed push button is a severity of 6 or 7 is not good value.

The bottom line is that they simply do not work. FMEAs are doodoo, junk, baloney and most of all bloody waste of time.

So as engineers how do we set out to "detect" the uselessness of this document?

-Test 1.
Towards the end of the FMEA add a completely bogus failure mode such as:

Failure: Fire due to short circuit
Cause: operator drops gold filling into assembly
Prevention: mandatory monthly dental inspection for all line workers

What percentage of this nonsense ever comes back to you for correction? Of the last 22 FMEAs that I wrote, the answer is 0, zilch! Obviously if I wrote this on any of the first 5 pages, I'd probably get called on it. But the reality is when you design a complex system, no one but the author will ever read all 68 pages of the document. Real world FMEAs are not team events they are a wretched design engineer toiling away late at night working to hit his deadline so some box can be ticked as "done."

-Test 2
Call 5 meetings titled "DFMEA >program name<" over the next year and track attendance.
At the same time call 5 meetings titled "prototype review >program name<".

Compare attendance and the quality of the staff in attendance. What you will find is that the experienced staff who can sniff an issue from a mile away will shun the FMEA meeting but will flock to the design review. Meanwhile, the bores and pedants in the quality department will always be present for your FMEA meetings and will be quick to question what statistical methods you used to determine the occurrence was a 6 and not a 7. Having experienced managers and peers is essential to reviewing a design in the context of failure mode prevention. Torturing them with 2 hours of mind imploding misery is no way to thank them for their time.

The back-door to a job in engineering

2011-11-20T18:59:00.001-08:00

Ask 100 kids what they want to become when they grow up and I'll bet you none of them will say " a human resource manager."

People with dreams and ambition don't strive to join HR, they simply end up there. HR people simply do the job. They hire as they're told, they fire as they're told and they dish out benefits and 10 year service award pens, as told.

So when a young smart ambitious engineer like you wants a job, who do you go to?

The only time you should ever deal with a HR department is if your qualifications perfectly match the job description. Then HR's job is easy: "yes this is what the engineering manager wanted, experience with FMEAs, understands GD&T, 10 years experience with auto electronics. Perfect." In which case it does not matter if you're a looser with the words "world piece" tattooed on your neck.

The problem with being an engineer straight-out-of-school is that your resume might not radiate with experience or some of the specifics the HR department is trying to match. Even though at university you were the brains behind a solar vehicle that crossed America, you had no formal training in junk paperwork like FMEAs, so you're sh*t outta luck.

So what do you do if you are a clever, bright, interesting engineer with a lot to offer but not a perfect fit to a job? Here's help from an insider:

Find out who the design manager responsible for posting the job is. How do you do that?

1) JFGI

2) Often email addresses are kept secret, so spend 30 seconds deducing it: name_surname@company.com

3) Verify your finds by pretending to sell V14GRA with your gmail address that you only use for anonymous use (and to the other design managers reading this, I know you're smiling and saying "so that's why I get so many of those emails").

4) Make the kill by sending lots of pictures or parts of the really cool stuff you've done and at the bottom of the box leave a copy of your resume. Easy.
5) Repeat 200 times.

Some of the best people I ever worked with, found us and not the other way round. So don't sit around waiting to be discovered.

Corruption, Extortion and Automotive test labs

2011-10-24T15:50:00.000-07:00

When I first entered the industry, I was perplexed by the question of why suppliers are tasked to perform their own testing. In other words, if I supply a window control module, then I am responsible for providing documentation that my module passes all tests. Until I produce this test report, I don't get my PPAP money. This is the equivalent to the Ivy League saying, "our profs don't have time to correct your exams, so go do it yourself and tell me your final grade"

No points for guessing that the 6th layer of Hell is dominated by test technicians and their managers.

There are two ways suppliers test their product:

Most major suppliers have their own test facilities. Therefore, the staff who determine whether a product "passes" or "fails" are paid by the same company that makes the product. So let's say it's a typical Friday afternoon and while you're running a radiated immunity test, you spot a slight twitch in the motor. Nothing too serious but the test plan clearly states "no momentary reaction." So you have 2 choices:

1) Not notice it and go home to a nice warm supper.

2) Take note of it and have the whole of the company's brass descend on your department, demand hourly updates and plan out how you're going to spend your next 3 weekends.

Smaller suppliers are compelled to outsource their testing. So let's say that it's the same typical Friday with the same motor twitch, only this time you work for an independent test lab. Your two choices are:

1) Fail the test. Send email to supplier telling them to come and collect the twitchy failed parts. You prance to your car knowing that you've created enough overtime income to take your two teenage daughters to see Justin Beiber live in concert.

2) Pass the test, knowing that if this problem ever happened in the field and resulted in a recall campaign, it's your head that everyone will be after.

This system, has another inherent flaw. Testing is very expensive. Passing tests is even more expensive. The right balance needs to be struck between running enough tests to capture the bulk of potential failures versus testing for every imaginable condition and failure mode on the face of the earth. It's easy to say "let's test for everything," except that when you do so, it is not only the cost of testing that increases. Now the cost of preventative measures for those one-in-a-million failures is tagged on each and every part. Extreme excellence is what car owners would like to own but not what they want to pay for.

So when a car manufacturer does not need to run component tests themselves, it's easy for them to start piling up test after test with extremely stringent requirements and often overlapping environmental conditions. In their world, they are simply striving for excellence. In their suppliers' world they are the customer to whom we quote 25% more in development costs.

Where are all the Women?

2011-09-15T20:58:00.000-07:00

It struck all of us engineers, the day we started college. As we walked into our first class surrounded by oil stained hands and stone temple pilots t-shirts we asked ourselves: Why are there mostly men?

Then we got our first job designing plastic electronic module boxes and we were shocked that amidst the faded cube walls, the engineering department had an even smaller percentage of women.

Why do women shun engineering?

Do you think this individual fantasizes about becoming a successful EMC test engineer?

I discussed this with an old professor friend of mine who has been teaching mathematics at University for close to 40 years and he observed that while at university, his female students tended to get better grades than their male counterparts. However there was a fundamental shift in attitude after graduation: Perhaps men are more motivated by the prospect of making more money and the boasting rights of inventing the next coolest gadget, and are ready to work harder for it. So after graduation there's a general motivational inversion that leads men to be more "successful." Of course, the choice of the word "successful" might be relative to your sex and perhaps if I was a 'Paula' Bearing, I'd be writing how women are more "successful" because they are much better at standardizing work, following procedures and ensuring that the workday ends precisely at 4:30 just in time to pick up the kids from daycare, cook dinner, put the kids in bed and have some quality time talking dining room color schemes with the husband. Many women who enter engineering don't see the fun of getting a call at 4pm from an angry customer who demands that they drop everything because one of the modules is deactivating when the horn is honked. Being the top engineer who is most gifted at figuring out issues with conducted emissions from a horn is not perceived as "success" to an average women engineer. On the other hand, a male engineer would find immense honor at being the customer's chosen one to "drop everything, panic and fix this issue." While in public, the engineer will curse his job and swear, in reality he would secretly much rather be honking horns all night long than read black beauty to his spoiled 11 year old brat.

I will end this blog post with a picture. Obviously, it's fake, but just in case you need help, here's why:

1) The chance of 3 good looking female engineering and 2 well kempt male engineers all in the same firm is as likely as Sara Palin joining PETA.

2) If all the engineers are smiling and wearing suits, it's probably because they have an interview for a new job later on in the day.

3) And why are they sitting so close together? Real engineers don't soap their armpits. This cannot be real.

4) What's up with the white blouse? Engineers eat their lunches alone at their desks and quickly learn that brown is the optimal color to hide salsa stains. Perhaps she's an intern.

5) And Ties? Have you ever tried soldering on a 0402 resistor with a tie on?

Why designers deserve to work long hours

2011-08-05T18:47:00.000-07:00

For the past 2 years, you've been working more than a fair share of weekends and late nights. There's often a couple of colleagues burning the midnight oil along with you. It always seems to be the same few. The irony is that you late-nighters are not the slow ones or the guys who've spent all day chatting. You're usually the most intelligent and thorough engineers. So why are you punished with perpetual late night work?

Here's the raw truth:

1) You work for a corporation that makes its money by squeezing pennies out of each process, each part...each person. You will be squeezed until you break. Grown men bawling ain't cool.

2) You snatched your last program from the jaws of disaster by using your incredible insight, experience and most of all, by donating countless of your personal hours to your corporation. Congratulations, you have just set the miracle as the standard.

3) You're the hero. You solved the last problem. Ohh by the way, there's another disaster for Mr Incredible to solve.

4) You've busted your ass perfecting a design. You could have released it as-is, but instead you squeezed in more testing, more prototypes and more reviews. It payed off because you spotted a couple of improvements that would have been to costly to fix after the program launched. You boss is ticked off because your DFMEA is rather shoddy.

5) Mike Thacher, your well spoken blue-eyed colleague from the launch from hell with the $500,000+ in engineering changes was treated by the company management to a post-launch basketball game from the VIP suite along with the CEO's hot daughter. Your ever so smooth program launch will get eclipsed by the 3 others that are following close behind.

6) What resource problem? The customer was pretty happy on your last project.

7) What lack of prototyping facilities issue? I just signed your Purchase Order for your new screwdriver set.

So Pawl's advice?
Learn to love angry customers. There's no better tool for getting what you need than a carefully managed angry customer.
-Stop hiding the truth. Let them find out that you are double timing them with another OEM.
-Make your customer jealous: "I don't know why you are reacting like this, my other customer told me what a brilliant idea it is."
-Tell your customer that they are growing fat and that they should cut down on junk documentation. You never know, it might get you divorced from a couple of programs. Then you can get back to your 1 hour discussions about how zippers were invented.

(step aside for the) Automotive Expert

2011-03-25T18:15:00.000-07:00

Today I was referred to as an "expert" in my field. After a split second sense of pride, I remembered that all my colleagues, customers and suppliers who co-share this title are twats.

Like teenage movie stars, technical experts start out as hard working, curious individuals. But once they hit "expert" status, self-glorification permeates their brain from one end while humility goes out from the other. Unfortunately, it is this very humility that makes us dig for deeper reasons and better ways of doing things. As long as there is a trace of doubt in what we do, there will be drive to improve. Eliminate this impetus and you've just killed what makes us become experts. Expertise is very much like the ephemeral life of mayflies where after years of life underwater, a few of the best ideas will emerge and triumph. But even the best of mayflies will become fish fodder after a day of two.

So when some young hard-working hot-shot named Bupesh comes along with a better way of doing things, the expert will thump his chest and roar that Bu is doing it all wrong and proceed to belittle him. He will misplace his anger towards his decline at the new guy and thus promote himself to official office wanker.

So please don't call me an expert. It might go to my head.

less is more?

2011-03-06T18:06:00.000-08:00

With the recession somewhat behind us, auto design offices are back in full swing. This time around, the products we design are a notch more complicated and the post Toyota pedal-gate era has left customers queasy and demanding about anything that contains a circuit board.

Only why are there still folks without a job?

With lingering bad taste of lay-offs, general managers are being cautious about hiring and only doing so in the absolute necessary cases. This means that these newer more complex designs that require more engineers, designers, lab techs etc are actually being handled by less people.

Source: Huffington Post

On the bright side, this forces us to be more efficient in what we do such as :
-relegate FMEAs deeper in to the "useless folder,"
-use more screen sharing software with our remote customers and suppliers and spend less time driving to meetings.
-Have more impromptu meetings with the team members that matter so that issues can be resolved as they arise vs in a monthly design review meetings.

On the dim side:
-There's no good way of putting lipstick on an overtaxed workload situation. The only way to avoid constantly working your weekends for free is to cut corners. Most of us have a pretty good idea of what design steps are junk (did I mention FMEAs?) and which ones are not. But there's always risk in doing so.
-Many GMs are playing brinkmanship with the human resource situation. When things go off the rails (people quit, baby-boomers retire or design sh*t finally hits the fan), there's just not enough momentum to keep the ship moving. Business gets pulled.

Customers are taking note of this and in their statement of work, no longer demand a list of roles working on their projects. Instead they are concerned with the actual names of the staff working on their designs. They are very quick to note that the same few engineers' names seem to pop up on every single list.

Perhaps that explains why all the projects assigned to me are for different customers...

Automotive word of the week - whiskers

2011-02-22T19:04:00.000-08:00

Prior to January 18th 2011, whiskers grew on your cat and face. Now they grow on intricate electronics in your car.
Tin or Sn whiskers are an odd growth where the atoms in stressed elemental metals migrate into slender needle like structures. You'd sooner figure out Gadaffi's logic than find a technical explanation for this phenomenon. Most notoriously occurring in tin, these needles can create electrical bridges across electrical components and the next thing you know, your ejector seat activates or worse still, your shaver goes dead when you've only shaved half your face.

While the occurrence of whiskers has been known for a while, the drive to eliminate lead from solder (typically substituted by silver) on your everyday circuit boards has led to a surge in this failure mode. Also, the minuturization of IC terminals means that these whiskers need to grow a lot less before they unleash flood gates.

Testing for this failure is not easy as whisker growth is heavily dependent on time; Tests lasting many years are unpractical; Thermal cycling circuits through extreme temperatures and humidity will trigger whisker growth, but the resulting whiskers are often too short and delicate to be noticed, let alone cause a failure.

January 18th 2011 was significant for Toyota because NASA cleared their cars of any electronics malfuntion. However they did discover a failure related to a whisker that bridged the redundant circuit to the main circuit causing both to go wonky.

When grouchy old engineers like myself design circuits, we usually specify a thin clear silicone coating that is applied over the circuit board so as to prevent this failure mode (and other more common failure modes as well). This process is known as conformal coating. Coating the connector areas is tricky because accidental over-spray on connectors often causes high voltage drops resulting in in-op circuits or fire. Given that fireballs have more dramatic RPNs than horns accidentally honking, we often leave the connector areas raw, thus making this part of the circuit most vulnerable to whisker growth. From everything that I see in the pictures of CTS' whisker prone pedal sensor device, it's not only the connector area that is unprotected, but the entire board.

Forget 5S and 8D, the new trend is 6F.

2011-01-13T18:50:00.000-08:00

I once walked into the service line at Walmart to return a broken blender while some looser ahead of me was returning a broken fishing rod. I was curious to what questions the young punk at the service counter would ask. I was expecting something like "hmmm I did not know that there were whales in lake Erie" or "do you own one of those bass boats that can troll at 200mph?" But no, punk boy asked for the receipt, yawned and gave him his money back.

Of course, punk boy does not care. He's just an underpaid employee. Walmart does not care because they will simply bill it back to fishing rods inc who in turn have no choice but to lump it.

Which brings us to today's topic: Warranty.

No matter how hard we try to protect our parts from being assembled upside-down and how many nights we spend tending to our 187 page long FMEAs, it's going to happen. Yes, just like facebook invites from schoolmates who we rather disliked, warranty will come knocking on our door.

These are the 6Fs of warranty (as applied to suppliers):

Step 1: Fightback (aka The denial).
When the customer first calls you to give you a "heads up" that the module is acting erratically, you brush it aside. You know that the customer's ECM programming has some bugs and that the guy who wrote the software was a victim of the recession restructuring. You sarcastically offer to help the customer debug the software.

Step 2: Fright.
After 3 days of software debug, the customer informs you that the problem only occurs on your parts which were made after November 13th. Ooops. You remember that the buyer switched suppliers of IC031 to some obscure Chinese company while the other supplier was on allocation.

Step 3: Fib
When your customer asks, you deny anything ever changing in your product.

Step 4: Flail (in futility)
Your customer is ticked off that you don't seem to be doing anything about the problem (little do they know how you are scrambling to restore the original IC031 supplier's parts back into the circuit and searching every last nook and cranny in the universe for the last reel of the remaining stock).
So you start running a whole series of futile root-cause finding tests.

Step 5: Finger-pointing
Customer blames supplier for inaction. Procurement department blames engineering for not forcasting the supply issue. Staff blames management for being understaffed...

Step 6: Forget
After 2 months of root causing and only 4 warranty claims, other bigger problems overshadow this one and your customer's engineer gets transfered to a new project. All is well.

The United States of China

2010-12-05T11:35:00.000-08:00

A few weeks ago, a Chinese electronics supplier called to say that they wanted to visit us. "How nice" I thought, these poor chaps are getting out of their mud huts and are blessed to visit America.
So clad in jeans and elbow worn sweaters, we welcomed the 3 Chinaman delegation. Although the boardroom's wall paper is from the 70s and grey around the light switches, it's functional. The carpet is well vacuumed and the anisotropic wear and coffee spill stains are subliminal. Nothing to make the Chinese feel uncomfortable.

When they did show up, we were faced with 3 well dressed young educated men who read for their degrees in recognized English Universities. They were accessorized in the latest technological gizmos and their powerpoints showed surgery ward levels of cleanliness and 5S in their shopfloors. Their accents were thick but no harder to understand that our colleagues in the southern states.

With their outstanding ppms on a few previous projects, they were very likely to secure the business. Unfortunately, they were outbid by their American competitors.

What's happening?

I won't even try to write down my thoughts when Thomas Friedman does such a great job in this article 'WikiChina' in the New York Times. I'm reproducing it for your convenience:

While secrets from WikiLeaks were splashed all over the American newspapers, I couldn’t help but wonder: What if China had a WikiLeaker and we could see what its embassy in Washington was reporting about America? I suspect the cable would read like this:

Washington Embassy, People’s Republic of China, to Ministry of Foreign Affairs Beijing, TOP SECRET/Subject: America today.

Things are going well here for China. America remains a deeply politically polarized country, which is certainly helpful for our goal of overtaking the U.S. as the world’s most powerful economy and nation. But we’re particularly optimistic because the Americans are polarized over all the wrong things.

There is a willful self-destructiveness in the air here as if America has all the time and money in the world for petty politics. They fight over things like — we are not making this up — how and where an airport security officer can touch them. They are fighting — we are happy to report — over the latest nuclear arms reduction treaty with Russia. It seems as if the Republicans are so interested in weakening President Obama that they are going to scuttle a treaty that would have fostered closer U.S.-Russian cooperation on issues like Iran. And since anything that brings Russia and America closer could end up isolating us, we are grateful to Senator Jon Kyl of Arizona for putting our interests ahead of America’s and blocking Senate ratification of the treaty. The ambassador has invited Senator Kyl and his wife for dinner at Mr. Kao’s Chinese restaurant to praise him for his steadfastness in protecting America’s (read: our) interests.

Americans just had what they call an “election.” Best we could tell it involved one congressman trying to raise more money than the other (all from businesses they are supposed to be regulating) so he could tell bigger lies on TV more often about the other guy before the other guy could do it to him. This leaves us relieved. It means America will do nothing serious to fix its structural problems: a ballooning deficit, declining educational performance, crumbling infrastructure and diminished immigration of new talent.

The ambassador recently took what the Americans call a fast train — the Acela — from Washington to New York City. Our bullet train from Beijing to Tianjin would have made the trip in 90 minutes. His took three hours — and it was on time! Along the way the ambassador used his cellphone to call his embassy office, and in one hour he experienced 12 dropped calls — again, we are not making this up. We have a joke in the embassy: “When someone calls you from China today it sounds like they are next door. And when someone calls you from next door in America, it sounds like they are calling from China!” Those of us who worked in China’s embassy in Zambia often note that Africa’s cellphone service was better than America’s.

But the Americans are oblivious. They travel abroad so rarely that they don’t see how far they are falling behind. Which is why we at the embassy find it funny that Americans are now fighting over how “exceptional” they are. Once again, we are not making this up. On the front page of The Washington Post on Monday there was an article noting that Republicans Sarah Palin and Mike Huckabee are denouncing Obama for denying “American exceptionalism.” The Americans have replaced working to be exceptional with talking about how exceptional they still are. They don’t seem to understand that you can’t declare yourself “exceptional,” only others can bestow that adjective upon you.

In foreign policy, we see no chance of Obama extricating U.S. forces from Afghanistan. He knows the Republicans will call him a wimp if he does, so America will keep hemorrhaging $190 million a day there. Therefore, America will lack the military means to challenge us anywhere else, particularly on North Korea, where our lunatic friends continue to yank America’s chain every six months so that the Americans have to come and beg us to calm things down. By the time the Americans do get out of Afghanistan, the Afghans will surely hate them so much that China’s mining companies already operating there should be able to buy up the rest of Afghanistan’s rare minerals.

Most of the Republicans just elected to Congress do not believe what their scientists tell them about man-made climate change. America’s politicians are mostly lawyers — not engineers or scientists like ours — so they’ll just say crazy things about science and nobody calls them on it. It’s good. It means they will not support any bill to spur clean energy innovation, which is central to our next five-year plan. And this ensures that our efforts to dominate the wind, solar, nuclear and electric car industries will not be challenged by America.

Finally, record numbers of U.S. high school students are now studying Chinese, which should guarantee us a steady supply of cheap labor that speaks our language here, as we use our $2.3 trillion in reserves to quietly buy up U.S. factories. In sum, things are going well for China in America.

Thank goodness the Americans can’t read our diplomatic cables.

Auto Designers: Sultans of snot

2010-11-19T15:43:00.000-08:00

Auto magazines like to use the title "designer" in the context of "Sang Yup Lee - The designer of the 2010 Camaro, moves to VW" or "Andrea Pininfarina, designer of Ferrari Enzo, Killed in Auto Accident." This leads the general public to think that there's some dude who stays late one night and armed with a red bull and a sketchpad, designs a car. Yeap, the man in black with a purple Karim Rashid chair has it all figured out.

Spring force on the handbrake release lever: Designed
Fuse ratings of each circuit: Designed.
Split lines for the bumper covers: Designed

In reality, all the designer produced was a cool manga sketch of a car that requires a periscope to drive and has such a tight gap between the wheels and fender that the maximum allowable suspension jounce is 2mm. The resulting turning radius compares with the QE2 and oops... it seems that someone forgot the mirrors.

Then come the engineers. They're not cool. They wear brown cardigans and have dull lives void of Can-Do attitude. They are focussed on finding fatalistic issues with every aspect of the design. They want nothing new, nothing different. Yuch. They act like Republicans.

Tension between pretty picture designers and engineering designers run high. Styling designers know that the exterior look of the car makes or breaks a car. If a design becomes a cult hit, they may become the next Philippe Stark of the car world. On the other hand, Engineering designers know that with a rear windscreen raked 3° from Horizontal, the car might become the next youtube sensation next time their mothers borrow their car.

7 tips for becoming a Quality Manager

2010-10-23T18:47:00.000-07:00

1) Difficult Situations: Grow Thick Skin
Replace the crushed apricot stone in your wife's ex-foliating lotion with crushed glass. After rubbing it all over your body, take a salt bath.

2) Impossible Situations: Practice Deceit
Tell your daughter "absolutely! when do you want to go?" when she wants to be taken to Disney world.
Tell your mother that you'll unblock her drain pipe as soon as you get back from your business trip.

3) Upset customer: Practice Apologising
Memorise Eliott Spitzer, Bill Clinton and the Pope's apologies and substitute "women" and "kids" for "missing diodes" and "broken snaps."

4) Act like the real thing:
Call yourself Bearing-San; eat with chopsticks and present your business card with 2 hands; 5S your desk and start by outlining a rectangular footprint around a picture frame of your firstborn. Label it "Kid #1."

5) Act like the real thing, Part 2:
Name your kids "Deming, Feema and Lexus "

6) Prevent loosing money on scrap
Feed the rotten apples to the dog (this is the automotive equivalent of using all the questionable parts for service parts [spare parts]).

7) Have lots of Sex and listen to really loud rock music...
Then genuinely claim that you can't see the defect and can't hear the rattle.

The differential. A must see video for all mechanical engineers...

2010-10-07T09:52:00.000-07:00

I wonder if Nokia invented their 'snakes' game after watching these synchronised motorcycle tricks. I also wonder if the production of this video led to the discovery of the crash helmet.

A day in the life of an automotive design engineer

2010-09-24T16:48:00.000-07:00

You get to the office 5 minutes late. You don't feel guilty. You were in till 11:00pm last night. There's 3 paper coffee cups and a red bull in the bin. For a position as lowly as the product engineer you are lucky to have a large desk. Looking at it, you know that you don't stand a chance of winning the top 5S work cell and the $20 'home depot' voucher that comes with it. The phone LED is red meaning that you have phone messages in your voice mailbox. Your inbox is also red: "253 unread messages." 120 of them are from yesterday and the rest are messages that still require action.

You are barely 3 minutes into reviewing the D599 latest circuit design when John, the new guy, stops by and asks you if you think that GM would buy-off on a 16 pin connector that does not have USCAR approval. You shrug. 5 minutes later you're back reviewing the circuit and you're worried that you'll never fit all those components on the 1 inch squared circuit board real estate left for you. You're also wondering how you'll ever dissipate so much heat from such a small area.

The phone rings. It's the buyer and he needs your help because the BCG423 is on allocation. There's an 18 week lead time to purchase the part caused by this post recession spring-back. We only purchase 18,000 a year so we don't have enough clout for the supplier to redirect the last of the stock to us."Pawl," can you find a substitute part? It takes 4 months to validate a change so why do people ask us that? In the end you know that you will be compelled to find a replacement and that all testing will be reduced to a single overnight stress test. The customer would never buy off on that so they won't be told.

So you switch gears to finding a substitute for BCG423. SGmicroelectronic makes a nearly identical one, but the power rating is slightly lower. You're nervous about the lower power rating so you speak to your boss about testing over the weekend. Unfortunately, the lab guys refuse to work the weekend so he tells you to get the chip supplier to do the testing for you. Last month we ordered the same supplier to lower all their pricing by 10%.

It's 9:30 am and time for your XT450 meeting in the 'Maui' meeting room. This business just got awarded to our organisation and our General Manager decided that this will be our showcase program. Rumor is that we were the only company that bid on it because everyone else said that the project timing made it impossible to design and test in time for start of production. Your boss wants a plan of how you will design it and test it in 5 months. You swallow hard. Midway through the discussions your mind rolls back into the circuit design you were reviewing. This circuit was due for production release yesterday. You also remembered that you need to book a hotel because tonight you're visiting a supplier. Then you drift off back into XT450 as the discussion turns to when you could have the first prototypes delivered. You say "3 months" but then you kick yourself because in the spirit of the moment you forgot to consider that the last 2 weeks of December are lost. You utter a silent curse.

The XT450 meeting gets interrupted by the purchasing manager who pops her head in to inform you of a mandatory attendance fire-fighting meeting about the BCG423 allocation. You were planning to check on the progress of some test on the ZD project. So you tell her that you'll try to be there in 10 minutes.

The ZD project is fairly complex and your colleague lost many nights of sleep designing it and dealing with the customer's indecision. During the peak, he was consistently working 70 hour weeks and often had to travel in the weekend. He quit 4 months ago and took a job as a calibration engineer with the government. Now you inherited the ZD. Your customer is pissed off at how slow your responses are. You told them that you'd have designs in 1 week. That was last week and Peter, your designer took emergency vacation.

It's noon. You like noon because people have lunch and leave you alone. You get back to reviewing the D599 circuit. You're getting increasingly nervous about the heat build up. You conclude that you need to prototype it and run some selected development tests. So you mail out the gerbers for quotation. You're impressed at how efficient these auto quote generators are. You're mid way into completing 3 page of the purchase order request form when the phone rings. It's the lab.

They inform you that "the ZD revision 12b stopped working during load dump testing". "Sh*t" you say to yourself, "the change did not work." You semi-jokingly call him an asshole. He chuckles. You try to forget about it and get back to writing the purchase order request form. The phone rings. It's your wife. "Honey, remember to pick up the kids tonight."

You write out a post-it note on your screen with the words 'Janet/John' to remind yourself not to forget them at the sitters again. You get back to your form.

It's 1 pm and the remnants of the smells from microwave dinners linger in the office. Now you're really hungry. Your screen flashes and a meeting reminder pops up:" Weekly engineers' meeting"

You take a bite off your apple and head off to meeting room 'Waikiki.'

This week your boss is stressing the need for you to record your hour-by-hour time allocation daily so that the projects can be billed appropriately. Also, your department scored miserably in a 5S audit with special remarks about "piles of dust collecting parts sitting on your desks." Finally, the president of the company is stepping out of the downtown headquarters and is visiting us. How nice of him. We are instructed to do a good clean up. You ask if it would be a good time to get the chipped 1970s bathroom tiles renovated.

At 2pm you interview a candidate who's fresh out of college. He said that he has no automotive experience but that he learned all about FMEAs at school and that he'd love to try them out on a new design. You don't tell him that his first job would be to rework misplaced resistors on 300 PCBs where our supplier missed the decimal point on our 0.4ohm resistor. He could start on Monday. You're wondering where you'd put him as there's no more desk space. You'll worry about that when he starts.

On your way back to you desk you stop and microwave your dinner. It's KD from last night (with frozen peas added after you heard a program on obesity on the radio). As you're waiting 2:20 minutes for it to warm up, new guy John asks you if the F23 plastic enclosure should be made out of nylon 66. "Absolutely not" you tell him as you recall last January's fiasco with all the tabs breaking off in dry weather.

You get back to your desk. Ketchup squirts on the D599 schematic. No one's looking so you lick it off. You must remember to bring a box of kleenex tomorrow.

The phone rings. The BCG423 allocation issue meeting has been postponed until 2:30 where your online calender is showing you as "available." The meeting is in the "Honolulu" board room.

On your way to the meeting you stop at the lab to see for yourself the issue with the ZD revision 12b. One of the caps seemed to have burst and the room smells of geek perfume (that's the smell emitted by microchips shortly after they make that "popping" sound). You start wondering what Taiichi Ohno would have done if he was in your position.

You stop for a bathroom break. You savour the serenity of having 87 seconds of thinking about nothing.

You get to the allocation meeting 4 minutes late. The general manager is there and asks you what you're doing to fix the supply problem. He uses the tone that implies "you idiot, why did you select a chip from NOsemi, did you not know they are having supply issues?" I point out that the part was designed before the recession.

At 3:45 you get back to your desk. The RED LED is still flashing on your phone. Messages must be piling up. As you glance at your sea of red email, one stands out. It is from Geoff Wilson, your customer, and the subject is URGENT. How original.

You open it. There's a new failure on the KD where your module is signalling an error to the BCM. He needs you down at the design center at 8:00am tomorrow. You don't dare say No.

So you start gathering your stuff to prepare for the meeting. You go to the lab to see if you can borrow the scope. It's out for calibration. You're trying to decide if you should book a hotel and drive there tonight or leave at 5:00am tomorrow.

Your wife calls to remind you to pick up the kids. She is proud when you tell her that you posted a sticky note on your computer.

As soon as you hang up, the phone rings again. This time it's from accounting who point out that you used the wrong cost center number for your purchase order request form and that you'd need to get the right number from the program director.

You're 5 minutes late for picking up the kids.

Your boss stops by your desk and points out that due to the recent surge in design related quality issues, the company is stressing the need for good FMEAS.

You reply "who is Pawl Bearing?"

This blog was brought to you by Tylenol extra strength. Take two with every missed deadline.

Benchmarking - It's hammer time

2010-09-08T19:17:00.000-07:00

In order to align ourselves with the latest styling trends and engineering innovations, automotive car manufacturer and suppliers often engage in benchmarking.

Benchmarking typically involves buying a competitor's product, analysing it's performance and then the fun bit starts: stripping it down. For us geeky engineers, this provides a nearly orgasmic experience as the products' layers are slowly and quasi-sensually peeled away. Ohhhs and ahhhhs follow.

A few years ago, when you could still use the words 'frills' and 'engineering' in the same sentence, I was invited to a mass benchmark session. This involved 5 brand new competitors' vehicles of the same class that were sliced, drawn and quartered. The entrails were then displayed on tables for experts in the different commodities to compare and be inspired with. There was also another of each of the same 5 vehicles that was left intact for engineers to be able to assess performance. I miss those days.

Nowadays when we want to benchmark a system, we just rent the car we want. Then, we are let down by the airbox that is blocking the module we want to probe. So engineers do what they have to do and remove a few screws and with a little forceful persuasion and a mallet, out comes the airbox. Invariably, under the airbox there's a plastic panel that still blocks our access but we can't remove it unless we disconnect the air hose. OK so you get the point. By the time we get to the module, there's a pile of 15 components, 58 screws and 4 springs next to the rental car. The performance of the module is never as obvious as we'd originally thaught so we close shop for the day. After a night of reading a brief history of time, our memory of the sequence of disassembly is warped, so the next day Alex argues "Nope Pawl, the bracket goes in after the cowl." Pawl replies "I don't think so and why are you still holding the spring that goes in before the booster bolt?"
Luckily, in this round of benchmarking, we manage to get the car back together and I don't think those 2 remaining springs were needed anyway.

Culture, class and corvette drivers

2010-08-26T19:42:00.000-07:00

Corvette drivers are not known for their subtlety or sophistication. I found this "toyota conquest" discount in a local 'vette ad very well marketed.

Opel Ampera styling cues

2010-08-14T18:23:00.000-07:00

Something always intrigued me about the styling of the new Opel Ampera (European brother of the Chevy Volt), though I struggled to place what that something was. Then this morning, as I was tending to the Japanese maples, the neighbourhood falcon squawked above me and it clicked...

Don't you think it looks like a falcon? After all, a falcon is not a bad choice of animals. At a diving speed of 200mph, it flies faster than anyone of us will ever drive.

The falcon in this picture is a Lanner. The picture was reproduced with the kind permission of photographer Grant Brummett.

10 ways to punish your supplier for being more profitable than you.

2010-08-10T04:03:00.000-07:00

Never mind that suppliers' employees work harder, get payed less and do things with common sense. Some OEM managers get angered when suppliers are more profitable than them so they find ways of paying them back.

Here are the key punishments:
-The day after reading their supplier's encouraging quarterly earnings,demand a $3M giveback.
-Force suppliers to write their own test plans.
-Force suppliers to seek approval of said test plan by 3rd world country engineers who have no clue what the product does and no interest how much it costs to test and don't give a sh*t that you don't have 3 months to get your test plan approved.
-Make them write a 60 page FMEA, then demand statistical justification for any occurrence less than 6
-Make them link every line of the DFMEA to the PFMEA.
-Route an air duct right through module, the day before design freeze.
-Have out-of-town suppliers attend weekly FMEA reviews every Friday afternoon.
-Make supplier change approvals so bureaucratic that it is impossible to improve a product.
-Create at least one unattainable requirement so as to delay suppliers' PPAP (=payment) for another 3 months.
-Send a quality expert who knows nothing about your product to assist you with preventing failures in your design.

Lemon Aid

2010-08-01T19:31:00.000-07:00

A few days ago, an article in the New York Times titled "G.M.’s Electric Lemon" really caught my eye. The blog belittles the Volt for being too expensive at $41k, too costly to develop, seats only 4 people, has little leg room and is the pillar that supports the entire GM organisation.

This article bugs me for the following reasons:

1) The term "lemon" is an after-the-fact condition. If someone feels cheated because they bought a car and their legs are not comfortable and they suddenly realised that the vehicle was developed with money that came out of their tax dollars, well that makes them an idiot as opposed to making the car a "lemon." Perhaps Niedermeyer is gifted with divination or perhaps whoever came up with the title needed to JFGI the meaning.

2) I always turned to the New York Times to learn. Lecturing the NYT feels so wrong.

3) Despite being a GM product, the Chevy Volt is Cool. Cool alone is a reason to spend $41k. This generation of the Volt is aimed at a privileged class and not your average Walmart-grey-Toyota-Echo. crowd. OK , OK so that Chevy Volt song and dance was not cool, but let's try and forget that.

4) The Chevy Volt is a pioneering technology = expensive, unreliable and highly susceptible to be out of date in a very short period of time. So even if it were priced at say $19k, it would be very unwise for the masses to be buying one at least for this generation of car.

5) Niedermeyer describes this car as an environmental "hair shirt." Now let's see, cycle to work, take the bus or drive the EV to work?

6) ..and he goes on and on about the $50b bailout funding and the energy department grants and tax rebates coming out of our pockets. Yes indeed, that sucks. But for that Money, I'd rather get a Volt than a retro styled Chevette.

7) In all probability, despite the skill of the engineering team and all the DFMEAs, there will be a significant hang-up. Maybe GM will overcome this hurdle or maybe they won't. If they do, they'll emerge stronger. If they don't, then go ahead and title your article "G.M.’s Electric Lemon"

8) "Only seats 4 persons." Well look around you and see how many people are riding to work in your car today.

9) The Volt may very well become the pillar on which GM is built. But if it were that easy to design a car that in its first generation became a ground breaking, market disrupting vehicle, then either your name is Steve Jobs or China would already be making them (actually a few Chinese companies are already selling EVs, but not quite a refined as the Volt).

10) I hate how pop culture denounces lemons. Think limoncella, lemon meringue pie. Think defence against older siblings when you squirt the skin at them. Or for Southerners like me who had the fortune of having a lemon tree on their property, think of the only citrus that didn't require pesticides to bear fruit, a tree that produced fruit regardless of the season. Think of the cure to scurvy and think of the fruit which can produce enough of a voltage drop to run a watch...but enough about volts.

Design Engineering: the real word

2010-07-29T16:19:00.000-07:00

Every time I interview a spring chicken that tells me how they dream of becoming a design engineer, it reminds me how 35 years ago, I had the same aspirations. I used to see myself as being the next Otto Diesel, Ma Jun or Elias Howe within a year of starting my new job designing square boxes for electronic modules.

Little do these engineers know that for the first couple of years they will be given the role of engineering janitors. In the years they spend rigging test harnesses and running wild goose chases on reducing the number of resistors used in products, they will learn to be humble in the face of the simplest mass produced part and appreciate how difficult it is make large volumes of highly reliably parts. Most of all, they will realise how only a small portion of their engineering time is actually spent inventing. Here's the break down of a typical automotive design engineer's work.

1) Modifying existing designs to prevent assembly operators from installing the circuit board the wrong way round, despite the nomenclature ink that reads "If you can read this, it's upside down."(20%)
2) Filling in forms such as SREAs, DFMEAs, 8Ds, Overtime requests, purchase approvals etc (20%)
3) Disposing of digikey cardboard packaging (5%)
4) Working on your boss' whim of the week (15%)
5) Working on your customer's whim of the week (22%)
6) Listening to your coworker's 1 hour daily rant of how busy he is (10%)
7) Looking for tools/parts that went missing from the shop floor (10%)
8) Driving to Ann Arbor, Marysville or Detroit (8%)
9) Designing/reviewing designs, packaging and calculating stuff -Aha finally (9%)
10) Convincing the storekeeper that you really need a roll of toilet paper now and that you don't give a sh*t that the union prevents him from working in his break (1%)

PS. If you're an observant engineer, you'll have noticed that 100% of your effort is not enough and that your working day is 10 hours long.

While USA is busy launching the new Ford explorer and abandoning cap and trade...

2010-07-26T20:04:00.000-07:00

...Beijing announced that it is set to begin domestic carbon trading programs during its 12th Five-Year Plan period (2011-2015) to help it meet its 2020 carbon intensity target.

With 17% of its electricity already drawn from renewables, this is another step that China is taking to gain energy independence and a much needed cleaner if warmer environment.

China's latest Economic Diode

2010-07-24T10:31:00.000-07:00

If you've ever worked on global cars like the Ford Focus or Honda Civic, then you're probably aware of how frustrating it is to have to break up a design and retool parts of the car with slight differences in order to meet the requirements of European, North American and Japanese markets. Since many of these requirements nucleated independently across the world many years ago, they are deeply entrenched in regulatory needs and local expectations. Over the years, there have been some small gains in the harmonisation of regulations, but generally, it's a hard thing to do. The most universally adopted requirements are the UNECE regulations so many countries outside Europe that lack the resources to write such specifications, simply adopt them. So if for example I'm selling 3 Bugattis in the United Arab Emirates, I don't need to smash 4 vehicles to ensure compliance with UAE standards on camel crash tests and crude oil and sand rub tests. I can simply use my European Type Approval. In return the UAE know that they're getting a safe car on their desert roads without needing to create and administer local standards.

So in 2003, when China decided to to launch their CCC standards, everyone swore. The good thing about the new standards was that they were cut and pasted from UNECE regulations, so it was a no-brainer to figure them out. However, the spirit of the regulations was left out of the "paste" operation. For example, one can influence the creation of the UNECE regulations by joining a series of working groups - Good luck trying to do that with CCC. Another example is that with UNECE, getting good parts certified is a relatively inexpensive and easy process and there are agencies spread in all the main manufacturing countries. With CCC you have a choice: Mainland china or nothing.

To make matters worse, the CCC authorities could not be bothered updating their regulations to be synchronised with UNECE regulation updates, thus resulting in subtle differences between the two.

Whenever you combine highly controlling, low paying govermental jobs with corporations in need of a signature, well you know what happens. Rumor has it* that there is a backdoor way to get your CCC approval in place.

Add all this together and what you get is a non-tariff trade barrier: Yet another economic diode that allows the flow of goods into China, a lot harder than those coming out of it.

*I can't confirm it's true but I've heard it from a couple of different sources and saw some fishy activity that supports this rumor.

Applying Automotive techniques to your spouse

2010-06-17T18:41:00.000-07:00

The trouble with my underwear was that the nice comfortable ones with pictures of bass and pike on them wore out really quickly while the plain Jane Walmart 6 pack-ones with itchy tags lingered on from my teenage days. The cause: bad FIFOing.

Once I learned the beauty of FIFO*, I started a regimen compelling myself to put the freshly laundered undies at the bottom of the drawer. Then, with military styled self discipline, I would only take from the very top (Though I'd allow myself an occasional deviation for a really important event like fishing with The Boys where underpanted hearts are an obvious No-No)

The challenge is now educating my wife to the benefits of FIFO in the kitchen, bathroom and bedroom. I'm failing to motivate her and she mockingly calls it "your FUFU," in that same sniggering voice I use to mock anyone who touts the effectiveness of FMEAs.

Perhaps I'll set aside FIFO for a while and introduce her to 5S.

*FIFO=First In First Out. The perfect demo of FIFO is Mcdonalds where freshly assembled big macs slide IN the chute in the order that they are made. The sales person naturally takes OUT the first made, hence FIFO. That way, they eliminate really cold soggy ones getting stuck in the system. The same technique is often used in Automotive parts manufacture to ensure that all sub components are of the latest pedigree.

GM's website: what's wrong with waste?

2010-06-01T19:03:00.000-07:00

I recently visited the GM website and I was welcomed by this picture. It immediately struck me as being wrong:

The picture shows two of the planet's most wasteful vehicles* in front of one of the planet's greenest way of producing energy. It's almost like the "new GM's" marketing guys were thinking "Ooh these young people like green stuff. So let's line up our cars next to some green technology." The sad thing about it is that GM are on track to being the first major automobile manufacture to launch an electric vehicle. A picture of the Volt with a wind farm as a backdrop would be a very symbiotic and symbolic combination.

*OK so the Yukon Denali in the picture might be a hybrid version making its consumption similar to an average car when driving in the city. The point is that these mega SUVs are the epitome of unnecessary waste, an image that jars with the backdrop of a wind turbine.

How to prevent Sh*t part design from happening

2010-04-06T19:40:00.000-07:00

Set Toyota's pedalgate scandal aside for a while.

When the book "the Toyota way" first revealed Toyota's 14 secret management principles that made the company successful, the biggest shock to all of us was that there was no magic elixir involved. It was all pretty obvious and boring stuff. Do the right thing the obvious way over and over again and hey presto, you'll be the best. Duh.
For example, principle 10 states Develop exceptional people and teams who follow your company's philosophy as opposed to the big 3 way of hiring some outsider CEO who might use the principles of selling lumber and screws to selling cars.
Principle 11 states: Respect your extended network of partners and suppliers by challenging them and helping them improve as opposed to dropping giveback bombs on your best suppliers.

Creating flawless designs is not much different. Doing it right shares the common theme of obvious and boring. These are Pawl's 10 principles on approaching a flawless design.

1) Hire bright and disciplined young engineers and keep them motivated and excited by means of best in class wages, constant training and respectable office decor. Constantly purge the bottom 10% performers.
2) There is only one document that is required: the test plan (DVP&R). When field failures happen, this document sets the line-in-the-sand as to who owns the problem and who will pay for fixing it. This does not mean that parties don't collaborate to prevent issues from happening, it simply makes problem resolution black on white and reduces the pressure on suppliers to sweep the dirt under the carpet.
3) The people who have vast experience and the ability to foresee a problem are also the busiest people. So when you invite them to your meetings, make your reviews effective by having CAD models or prototypes ready for them to feel and touch. Never ever waste their time with a FMEA.
4) You can never spend enough money on prototypes.
5) Your designs are only as good as they can be built. The chances are that your design is being assembled by the bottom of the barrel people or by bright people who cannot speak a word of English. So when you review the process flow with the manufacturing engineer and he complains how easy it is to assemble something upside-down, don't poo-poo him. Just fix it.
6) Retain in-house know-how of all your outsourced designs otherwise you'll be screwed.
7) No matter how much you trust your supplier, create a statement of requirements or drawing for everything you purchase (Again, the line in the sand of who needs to do what and by when).
8) As hard as you might try to achieve a perfect design, problems (just like sh$t) happen. When they do and you learn a lesson, reward your counterparts with some shadenfreud and share the lesson with them.
9) Don't try to reinvent every wheel. Standardise commonly used parts and commonly used approaches. Reward customers who use your standard parts by means of financial incentives.
10) Your design engineer did a great job over and over again. If you reward him by dumping an excessive number of projects and problems on him, his efforts will be so diluted that he too will do a sh*t job. This greed is the number one overlooked failure mode. Go ahead. List it in your FMEA.

Pawl's call to war on FMEAs

2010-03-29T18:47:00.000-07:00

As a servant of the deity of design efficiency, I order you to fight the unbelievers of practicality, to slay the followers of futility and to reign victorious over the FMEA faction. The fight will be long, but my brothers and sisters, you will be rewarded with the supreme knowledge of usefulness and the eternal glory of the almighty zero defect car and truck.

These are my 4 simple steps in crushing the FMEA enemy.

My loyal allies, you will start off with minor disruptions to the process by bringing the document into disrepute. Examples of this are: Passing wind during a meeting; Or, sticking to the same theme, listing irreverent failure modes such as "Hydrogen Sulphide emitted from the driver's posterior orifice, reacts with the silver plating on the IC lead-frame causing the gold wire bond to corrode prematurely resulting in false activation of the steering lock. Severity 10."

As you become more comfortable with your assignment, proceed to more sinister disruptions such as modifying the master FMEA excel spreadsheet so that the RPN formula is given by (severity X occurrence X detection X RANDOM{1,3}).

Congratulations, now you are an enlightenment green belt. Your next task is to copy Philip Morris' techniques in casting doubt behind the science behind FMEAs. Start off with Wikipedia and modify the entry to reflect the growing scepticism of the FMEA's usefulness in the automotive engineering world.

As the newly crowed Sultan of truth, you are ready for the kill. As a warrior, you are ready to perform the ultimate sacrifice.
1) Infiltrate the U.S. National Highway Traffic Safety Administration and become the CEO
2) Next time there is a recall send in the FBI to seize the auto manufacturer's PPAP binders
3) Bring up the DFMEA and expose that despite having an RPN of 288, they deemed it fit that "current controls are adequate. No further action required"
4) Chuckle as the vehicle manufacturer issues a secret decree that on all future projects 2 DFMEAs will be written: The one on white paper is the "NHTSA" version and the one on green paper is the "true version"
5) Chuckle again as the vehicle manufacturer comes to its senses and say "F*ck it. We're not doing FMEAs any more."

Eternal salvation is yours Pawl.

The truth about automotive service (spare) parts

2010-03-25T17:43:00.000-07:00

Last time you paid $500 for a part as trivial as a replacement plastic bumper cover, it probably left you wondering if you'd just been ripped off. Have you?

Typically, service parts cost 10 times as much as the vehicle manufacturer pays for them, so in reality under a mass production environment, the manufacturer is only paying $50 dollars for your $500 part. The bastards. Well yes and no. Your $500 dollar part required special non returnable packaging. Also, there are six different colors for both high end and low end bumper covers, so that's lots of big parts that need to be stored in many warehouses. Under a mass production environment, these parts are sequenced so that they come into the car plant in exactly the same color and order as which the cars are being built. This way there's no storage cost at the plant. The dealer who brought the part in for you made a decent cut on it. So did the transport company that delivered it. If the part is no longer mass produced, the small volume runs have very expensive assembly line start-up costs associated with making the part. When you add up these costs, there is a huge cost increase to the part.

On the other hand, the manufacturer knew that you have no other choice but to buy their part so unlike a built up car, there's little competition.
Finally, and this is the fun bit, the OEM supplier (such as TRW, Visteon, Delphi, Magna etc) knew that there's no quality inspectors looking at service parts' quality so instead of disposing their fish eye, orange peel and scratched paint rejects in landfill, they sold them to the vehicle manufacturer as service parts. The chances are that your car had a fine coat of road dust on it when to took it in for repair so you'd never notice a few superficial defects. Since most of the borderline stuff gets assigned as service parts, generally spare parts are consistently worse quality than regular production stuff.

OEMs are typically contracted to manufacture service parts for 10 years after the last production part rolls off the line. Once again, the number 10 naturally falls in place because it coincides with the average lifespan of a car. After 10 years, the market is flooded with spare parts from retired vehicles so no one in their right mind would pay premium price for a spare part on their old banger.

And that is the truth about spare parts.

Program Managers: The linchpins of the automotive development world

2010-03-16T17:46:00.000-07:00

Most automotive manufacturing companies are organised into stand alone departments like quality, design, sales, operations etc.

When a new product gets awarded to the company, the various departments are all out to protect their group's interests. This sometimes leads to tricky situations:

For example: We are awarded the 2013 model year XV269 BCM; Volumes 125k/annum.
Sales manager wanted the project at all costs and under-quoted. Now quality wants a 48 hour burn in instead of the usual 10 hour. Manufacturing say the module is too small and that we need to grow the board by 20%. The design engineer is being pressured by the customer to reduce the size by 40% as the BCM is interfering with the cigar holder. Purchasing want to try a new Indian IC manufacturer who just finished developing a knock off ST instruments driver that is 60% cheaper. Accounts will not let us purchase the drivers 10 weeks in advance, even though they are under 'allocation' because of the sudden resurge in orders due to a recovering economy.

OK OK, you've been there done it, everyday office politics.

This is where the program manager comes in.

The program managers are answerable to the product and not to the department so they act as crosslinks between the departmental 'mers' or as the title suggests, the linchpins that keep the wheels on the axels. They have the ability to call the department manager's bluff and they tend to act as the ombudsman for intra departmental litigation.

Being a program manager is a pretty prestigious job but before you rush off to become one, keep in mind that:

Most automotive companies' salary is proportional to the number of people reporting to you. Typically no one reports to a program manager.
The expectations are very high while the authority is somewhat limited especially since the people doing the work don't actually report to the program manager.
As a result, many program managers burn out way before retirement.
For those that do survive this trial by fire, the exposure to the full spectrum of the organisation makes them great candidates for the next General Manager, especially if they're blue eyed boys.

How to become the top dog of your engineering office

2010-02-24T18:31:00.000-08:00

Grandpa, what have you done all your life?

Boy, I studied all there is to know about engineering and then I used it to develop automobile bumper covers.

Pretty cool huh?

If you don't think so, here's 10 easy tasks to reach the highest echelons of an engineering company:

1) Say yes, yes, yes for 500 times. Now you're prepared for the next time the CEO requires something of you.
2) Be a blue eyed boy who excelled in college sports
3) Count how many engineers you know who became heroes by preventing ill conceived projects from happening. Heros solve problems, they don't prevent them. Now say "yes we can do it" for 100 times.
4) Who pays you? Instead of focusing your efforts on pleasing your customer, try pleasing your boss.
5) A perfect product launch is a perfect reason to give you more work; a problem riddled launch is a golden opportunity to shine.
6) If you ever read a book that starts with "Howard Roark laughed."- Burn it.
7) Your boss is 10 times more likely to get ticked off at you for saying "I don't have the time to do it" than if you said "I did not have the time to do it."
8) Ask yourself "what does my boss gain from promoting me?" If your answer is "nothing, he would lose his star employee," congratulations! You've just figured out why the most competent engineers are stuck in the same job forever.
Corollary: The way up is a zigzag not a straight line.
9) Never Panic. Say "it's only a car" for 20 times.
10) During question time with your CEO, which of the following 2 questions will earn you more points: "How come we made $1.2 billion in profit and yet we got no raise?" or "Since you came on board, we made $1.2 billion. Is this due to your exceptional management skills or is it due to your extraordinary vision?" Now lick your chocolate ice-cream 10 times.

Barbie computer engineer

2010-02-21T07:36:00.000-08:00

Now if only they'd make a barbie quality engineer. I'd buy one and stick pins into it.

Allocation: Why this purchasing hell-word reflects a recovering economy

2010-02-19T20:12:00.000-08:00

There's no word that rivals "allocation" in sending shivers down a buyer's spine.

Allocation is a common term used in the purchasing of electronics that essentially means "demand is greater than supply" and there's no way you'll get all the parts that you need in the time frame that was promised to you. As a buyer you are forced to pull all your "allocation" tricks like threaten to pull the business away or worse still threaten to send your quality manager to assist with allocating the parts to no one but you. Pain.

However it also means that the economy is back on track and that the purchasing world is back to normal. Hooray, for allocation, it's the first sign that the recession is over!

Dr. Seuss' advice to Toyota

2010-02-15T17:24:00.000-08:00

Extract from "Oh! The places you'll go!"

You'll be on your way up!
You'll be seeing great sights!
You'll join the high fliers
who soar to high heights

You won't lag behind, because you'll have the speed.
You'll pass the whole gang and you'll soon take the lead.
Wherever you fly, you'll be the best of the best.
Wherever you go, you will top all the rest.

Except when you don't
Because, sometimes, you won't.

I'm sorry to say so
but, sadly, it's true
that Bang-ups
and Hang-ups
can happen to you.

You can get all hung up
in a prickle-ly perch.
And your gang will fly on.
You'll be left in a Lurch.

You'll come down from the Lurch
with an unpleasent bump.
And the chances are, then,
that you'll be in a Slump.

And when you're in a Slump.
you're not in for much fun.
Un-slumping yourself
is not easily done....

However the Doctor goes on to explain:

...and remember that Life's
a Great Balancing Act.
Just never forget to be dexterous and deft.
And never mix up your right foot (pedal?) with your left

And will you succeed?
Yes! You will, indeed!
(98 and 3/4 per cent guaranteed)

Lincoln MKQ and A

2010-02-12T19:14:00.000-08:00

If Lincoln's sales team studied the demographics of people that buy their cars, they would have quickly realized that the names they give them should be logical and easy to remember. You'd think that they would have come up with the Lincoln Zenile or Lincoln Pearly Gates. But no, the young punks played a joke on old grandpa and made all the cars look and sound the same.

So just before pops goes off to buy himself a new car, give him this test sheet and make sure that he comes home with the right car.

Engineering: The good enough profession?

2010-02-07T16:05:00.000-08:00

Bankers are quick to learn that there is no better way to instill a sense of confidence than by looking fresh and smart: A well known bank chose the shade of blue in their cubicle walls to match the blue in their logo. After five years of use and a couple of coffee stains, these cubes were replaced with something trendier. Likewise, corporate employees of many banks are reminded that a suit lasts 20 uses, no more. Even if you still smell traces of store newness in it, give it to the Salvation army and for heavens sake buy a new one.

Our engineering firm picked these cube walls up for a really good deal. Who in their right mind would sell such good cubicles for that price? A good dusting and they were as good as new. The shade of blue even matched the burgundy ones that we already had. Everyone wanted one of the new cubicles. Some of them still had stationary in their drawers. Mine belonged to Mindy Hawkins, investment adviser. I don't know what she looked like but she smelled nice and was the well prepared sort with a spare pair of panty hose in the second drawer and 4 spare toothbrushes in the top. She left $4.56 is spare change.

We settled into our new sky blue cubes and pinned up pics of the family on one hand and timelines on the other. Then we started engineering the new product. It was a very complicated and expensive set of parts that created a personalized audio-visual sensory stimulation in the car's cabin. Not any car, this was a really high end car. I was so lucky that I even got to ride in one at the proving grounds. We're going to make some banker or lawyer so proud. Tonight, instead of changing the oil in my '98 intrepid, I'll work late to finish off the final details of the design. Perhaps Mindy will buy one. It will be my little way of thanking her for leaving me enough money to buy a few "A8" coffees from our beverage machine.

Engineering: Profession of a lesser god?

2010-02-04T17:54:00.000-08:00

Garnering respect for one's profession does not come automatically. It comes from generations of collectively succeeding in very challenging tasks. Engineers do a great job at this. However respect for one's profession also involves a degree of self-respect. All too often, engineers are comfortable showing up to work in casual clothes or worse still, they are tolerant to working in an office reminiscent of a 1980's communist shipyard drafting-room.

Here's a list of stuff that that you ought to do for your profession:

1) Dress well. No need for suits, but whatever it is, it should be under 1 year old and when granny offers to patch the elbows, tell her she can keep the sweater for her cat's litter.
2) Do not be content with a crummy environment. You're a professional for heaven's sake. This is what the office of a real professional looks like.
http://www.officedesignblog.com/
3) Speak eloquently and refrain from using foul language, unless of course you're talking about FMEAs.
4) Have zero tolerance to shoulder dandruff on anyone who calls themselves an engineer.
5) Restrict your urge to smearing your body in used engine oil to your project car not your work car.
6) Be aware, knowledgeable and respectful about professions outside engineering and don't poo poo them (and yes it was wrong of me to call dental hygienists - glorified toilet cleaners).
7) When overworked, don't dismiss the importance of the jobs being asked of you. Simply explain that you are very busy and that you are working on prioritizing the tasks ahead of you. Only then, give in to the urge to hit the delete button. This instantly transforms the perception of you from ar*e hole to in-demand professional.

8) Face the fact that how-good-your-product-looks is on par with how well it works. The folks in the styling department might very well be self-centered prats, but they probably have more effect on how well a car sells than you do.

9) Read Henry Ford's book Today and Tomorrow over and over again.

10) Even though the problem that you're facing has an RPN of 597, you are not exempt from saying a good morning hello with a smile on your face.

The dental business

2010-01-31T19:22:00.000-08:00

I work for one of the most profitable OEMs around. I make a decent salary and apart from FMEAs, I enjoy my work. Life is bliss until I go to the dental hygienist. Every 9 months a young girl with 2 years of community college training scales my teeth and swooshes them with strawberry flouride. To while away the time, I calculate what a reasonable bill should be and come up with a figure of $120. When she's done flossing, she bills my insurance $180 and gives me a toothbrush.

At first, I'm happy to see such a high bill because my insurance pays for it and I look at it as $180 avoided, but deep inside there's that nagging feeling that:
a) This glorified toilet cleaner must make a hell of a lot more money than me.
b) Although the insurance company is being routinely bilked, it's your employer that's ultimately footing the bill. There's no incentive to shop around for better priced dentists and hygienists so you choose the one with the fancy receptionists and TVs.
Think of all the engineering candy that your employer could have bought with the (excess) money spent on dental care: a fancy SLA machine, new fast computers for everyone or maybe even an orgasmic $100 digikey certificate that could be spent for work projects without managerial approval.

PPAP smear campaign

2010-01-27T18:15:00.000-08:00

In the dreamy world of quality engineering, PPAP (pronounced "pee" as in urinate and "PAP" as in the cancer test) means "Production Part Approval Process" or in simple terms, how to get a part approved for use on a mass produced car.

In the real world of automotive engineering it is a noun or verb generally preceded by a profanity. This is how it works:

The vehicle manufacturer awards you a project and a contract is signed. You pay your product development department hundreds of thousands of dollars to develop the product and when the design is ready, you award your suppliers contracts for tools, fixtures, assembly lines etc.
Until now, not a dime has been paid by the vehicle manufacturer.
9 months after your award of contract, your suppliers deliver the equipment and you make your first-off parts, assemble them and test them. Like all other projects, you identify a few issues and make some costly changes.
After 14 months and 5 million dollars worth of investment, not a penny has been received.

This is where the PPAP (noun) comes in:
When you PPAP (verb) a part, you are sending your auto manufacturer a massive document showing compliance to a whole slew of test requirements, completed forms, checklists, drawings etc.

It's in your interest to get PPAPed (verb) as soon as possible. That way, 1.5 years after emptying your entire bank account, you're finally seeing the money come back. You'll be able to pay the tool shops; The kids in Windsor, Ontario will no longer go hungry.

Well, it's not that easy. The vehicle manufacturer knows that as soon as they pay up, they instantly give up all incentive for you to play fetch or do whatever mediocre task they give to you. And the compound interest of 5 million dollars is a lot of coin so they're in no rush to sign your PPAP.

And if your projects are ever like mine, there is always some silly little EMC test at 200 kHz that you just can't pass. Now your customer is suddenly seriously concerned that if a driver adjusts his seat while driving and listening to the AM radio in Africa, maybe, just maybe he might hear a slight hum. "Sorry Pawl, we can't give you a deviation."

Situations like this led to the creation of a new type of PPAP: The Interim PPAP (noun) is like the 1st layer of Hell where all the great non-believing philosophers and un-baptized kids are sent. This limbo land is for suppliers who are missing completion of a trivial test or two. Good enough to ship saleable parts but not good enough to get paid for tooling.

Perhaps PPAPs should be renamed as "Piss Poor Attitude to Payment"

As Toyota's quality tends towards infinity...

2010-01-22T19:46:00.000-08:00

... you will never need to buy another car again.

Applying theoretical limits sometimes reveals trends in more subtle events. So while everyone is quick to blame recession and floor mats as the reason for Toyota's dwindling sales, not too many gurus are applying the lessons that we learned from Steven Levitt/Steve Dubner in Freakonomics: In this book, the authors pick out remarkable trends in a diverse range of fields that provide answers to previously misdiagnosed causes. Their most notable example was the demonstration that Roe vs. Wade had a ripple effect on crime 16 years later because many babies which were predisposed to become criminals were never born.

What I'm getting at is that Toyota's dwindling sales is in part due to the remarkable quality of their cars manufactured 12 years ago. As Toyota's ad recently reminded us "80% of all Toyotas sold in the last 20 years are still on the road." So if my Toyota now lasts 12 years instead of 10, then currently I will purchase 20% less car. As the vehicle expected lifetime becomes longer, the necessity to replace a new car diminishes. Now some of this is blurred by shifting market share, recession, GDP etc but the underlying need to replace your car less often still remains. In the past, Toyota's sales decrease due to exceptional quality was offset by their dramatic increase in market share. However there is a limit to how much of former big 3 business you can lure. As the market share tends towards 100%, then the exceptional quality issue starts becoming more and more of problem. Who would have guessed that "quality problem" would one day have a counter intuitive meaning of too much quality?

So next time you get a box of warranty parts on your desk don't be disappointed in yourself. Keep on doing what you're doing. In the long run, keeping that ppm above the zero mark will help your customer survive.

Confessions of an auto-design interviewer

2010-01-16T20:25:00.000-08:00

In the good old days of the auto industry, our HR team hired a bunch of idiot designers and engineers to help me get things done. I suspected that they were dumbos after spending about 30 seconds with them. All the signs were there: 3 missed beard hairs on their first day of work and repeat occurrence of the same elbow worn sweater after only 4 days of work. After 1 week, one guy boasted that during the last election he voted for candidates based on how sexy their wives were. 2 weeks into the first assignment I knew these guys were lemons and the only thing that they were good at was tapping into every single benefit available.

I quickly learned not to trust our HR department. They reminded me of realtors and filling in the vacancy as quickly as possible seemed more important to them than finding the right person.

After becoming an interviewer myself, I came up with a few trustworthy observations. Here's the list:

1) Count the number of grammatical errors and typos in the resume and multiply it by $100,000 dollars. This is the yearly cost of dealing with the errors made by this fellow.

2) Look hard at the interviewee's hand. If there are traces of engine oil under the fingernails or missing finger tips give him (or in extremely rare cases 'her') some extra marks. Hands-on, just the way we like them.

3) Ask them what they think about FMEAs. If they have more than 3 years' experience designing stuff and use the cliché "a useful tool to prevent failure modes," tell them to jerk-off. By the way, the perfect response to this question during an interview is "after many years of writing FMEAs I struggle to appreciate their usefulness. However I understand how many of your customers hold this piece of paper very close to their hearts and I will ensure that my FMEAs will not be the detail that holds up your PPAPs."

4) If they have a PhD in engineering from some Chinese university, ask them to draw a stress/strain graph for any metal they like. If after 30 seconds you don't get an answer, draw it for them and ask them if they like your picture of Mount Everest.

5) There might be instances that could catch you completely off guard. In one interview I gave, a fellow who had been unemployed for over a year and who wore a bright blue jacket, gave such a bad answer that I felt a little laugh coming out. I managed to suppress it by turning my head away from blue-boy and instead looked at my co-interviewer. Geoff. The trouble is that Geoff was starting to turn purple as he endured the pain of holding in a giggle. Then we both burst. Since then I learned a trick: If you feel a laugh coming up, bite your lower lip really hard and for heavens sake, stare at nothing but the floor.

6) Show the interviewee a component that failed in the field and ask them why they think it failed and what they would do to fix it. Rule 5 above may come in very handy.

7) Never ever ask the question "what is your greatest strength." This cliché is equivalent to playing the opening riff of 'stairway to heaven' in a guitar shop and is reserved for interviewing candidates at Wal-mart.

The big bad supplier is gonna get you

2010-01-11T20:51:00.000-08:00

To all the suppliers teetering on the brink of extinction, this recession might be the last nail in the coffin. You're doing your best to keep afloat but cash seems to flow out way too easily and deep inside you know it's a lost cause.

If you're working for one of these anemic corporations, you might have noticed strange new faces around with excel spreadsheets of every tool in your factory. How do they know all this stuff ? Why are they so interested in the hole pattern in all our dies? Why do our managers never smile at them or offer them coffee? Who are these people?

They are dementors coming to suck out what's left of you: all the injection molds, assembly lines tools and wave-solder carriers. They stop short of taking your capital equipment like your pick and place machines and your progressive stamping machines. No, they'll let you go belly-up before picking this lot up for a song.

Welcome to take-over time.

Plagiarism, Extortion and CCC certification

2010-01-06T19:09:00.000-08:00

Just as you thought it was hard enough to create global products that simultaneously conform to MVSS, UNECE and JASIC requirements, in come the Chinese and say "evly supa powa has specification. So we make oul own."
You bang your head against your desk and get down to studying the new GB spec.
Barely one paragraph into it you realise that it sounds familiar. Very familiar.
In fact it is so familiar you can continue reading it with your eyes closed.
This is because when China "wrote" their specs, they did what they do best: They copied it word for word from the UNECE regulations.
However being a communist country, they decided that the ECE system was to unbeurocratic for their likings.
So they complicated it.

Here's how it's done for ECE:
Call up a type approval house and set up an appointment for them to witness the testing. 2 weeks later get certificate and invoice.

Here's how it's done for CCC:
Wake up at 4am to call up the CCC Authorized Certification Body (ACB).
Find they only speak Chinese
Hire a Chinese translator
Find the ACB is not authorised by CCC to assess your product
Call up another ACB
Fill in the CCC application
Discover that there are many unwritten rules
Resubmit CCC application
Get certification.
No, no it ain't over...
Send parts to CCC approved lab. All the approved ones are in mainland China.
Re-hire translator to tell lab how your stuff works
Get certificate
No, no it ain't over...
Now your factory must be approved
Buy plane ticket for Chinese inspector (The only official inspectors all live in mainland China)
Book and pay for their hotel
Welcome inspector and tour them around factory (beware: they might steal all your ideas)
Pay the fees and yippee you've got your CCC!

Ohh and renew application every f*cking year.

5 reasons to love your design engineering job

2009-12-31T10:47:00.000-08:00

My first task as a fresh-out-of-school design engineer was to develop a plastic slider part for an electronic mechanism. Although it was 90% copied from existing product, it took me 3 months and I was extremely proud of it. This slider was the only thing I could think about. One night, as I was having a discussion with my girlfriend, I drifted asleep. She asked me some question and in my dormant state I replied "slider."

It was at the age where all my college friends were discovering the joys of having a real job. So when I finally received the first few molded sliders, I took them out to the pub with me to show them off. The reaction was not what I expected: In my mind, I was holding an integral piece of equipment that would help transport countless people around the world; the culmination of 4 years of complex mechanics and calculus all summed up into a single part. In their mind it was a piece of plastic.

Here's some of the moments that we, design engineers, live for:

1) After staring at a part in 3D for many months, we finally see the first real thing.
Note: A funny feeling often accompanies this moment called 'size psychosis' where you realise that all along you have been imagining things to be 10% larger than what they truely are.

2) Seeing the car you've been working on at the auto show. Or if you're fortunate enough to work for Boeing, watch the dreamliner take off.

3) It often does not take much to root cause a problem that results from a common failure mode. However when it takes two or more factors to cause a problem, things start to become pretty mind boggling. The 'Eureka' moment when you finally figure out the problem really makes your day.

4) There are a lot of stake holders involved in launching a product and we often do so with conflicting objectives. Many times, despite our warnings of Armageddon, design engineers' directions get quashed by senior managers or buyers. 6 months later, the "I told you so" goes down so sweet!

5) Having all the ladies drool all over you because you're a design engineer (OK OK dream on Pawl)

Love thy Engineering Change

2009-12-27T10:03:00.000-08:00

Automotive design engineers dream of the day when we'll release our designs for manufacture and never hear about the product again until we receive a letter from the president of the OEM thanking us for our seamless design.

Not gonna happen.

Instead on Friday afternoon, one day after tooling release, the phone rings. It's your customer:
"Pawl, the air cooler guys changed their routing and we lost our 5mm clearance"
"But I checked it in Teamcenter and it's fine"
"Sorry, Pawl, they didn't have time to databank the change"
After an eerie 3 second silence:
"Pawl, we need you to fix it."

Any self respecting engineer would utter a couple of obscenities, remind himself to find a new job and then get back to work in a really bad mood. Right?

Wrong. As long as this change is not a result of your oversight, engineering changes (aka EWOs, change notices, change orders) are your best friend. While not being very illustrious, they are a licence for your employer to charge obscene amounts of money for overinflated tool modifications, tool shipping, obsolescence, expedited engineering, testing, administration, re-PPAP. Hahahaaaa >insert evil laugh here<

Obviously, since you work for a highly profitable organisation, the expectation is that like a good engineer, you will donate your weekend and really impress the customer at how quickly "we" can solve their issues. After all, you're just an engineer and you didn't have much to do this weekend besides fix the WC.

However, your program manager steps in and is charging the customer $120/hour for 25 engineering hours for the 10 hours of work that you are donating this weekend. No fair.

Figure 1: Loving thy Engineering Change classic example: Note how the dinghy is called "original contract"

Bad way of dealing with the situation:
Make up an excuse and say that you will not be able to get to the job for another week. In the process you'll tick off your customer and your employer.

Good way of dealing with the situation:
Step 1: Tell the customer that it's a really complicated job but you're sure that the whole design office is available for work. Tell him to call your boss.
Step 2: Immediately call your boss and ask him if it's OK to leave 1/2 hr early today because you are going on a much needed family weekend break.
Step 3: Sit back and wait for your boss to appear on his knees
Step 4: Ask your boss whether it would be fair to balance the let-down with having the company pay for the family weekend break.
Step 5: Complement your boss on his wise decision making skills.
Now everyone is happy.

10 signs that you should quit your auto job

2009-12-17T19:41:00.000-08:00

1) You give your life and soul for the success of the billion dollar company you work for. You are rewarded with business cards printed on DIY Avery business card paper. You detect a slight snigger on your Japanese supplier's face as his fingers run along the micro perforations.

2) You are trusted with releasing the design of a 5 million dollar project but if you need a pen, you're sh*t out of luck because they are kept under lock and key.

3) Your new plant manager was brought in from another industry one week after your general manager announced that the company will be adopting TPS philosophy.

4) You are expected to make the 5 hour drive to Detroit after a 5pm meeting. You are not allowed to expense coffee. AKA corporate downsizing.

5) You are transferred to the quality department

6) 5 or more of the following things are missing from your office: natural light, picture frames with up-to-date photos of the vehicles your parts go on, live plants, soffits with no water stains, a carpet that is newer than 10 years old, free coffee, a secretary.

7) Getting a PO of under $250 approved involves more than 2 approvals

8) Getting a PO for under $250 takes more than 24 hours

9) Your gift for 25 years of service still has a dollar store label attached

10) Your computer has a 'Y2K compliant' sticker on it.

Farewell to art

2009-12-16T14:28:00.000-08:00

I thought that the Volt was cool until I saw this abomination on Auto Prophet's blog

Perhaps GM need to extend FMEAs to their marketing team:
Failure Mode: People won't buy our car
Effect: We won't repay the people
Cause1: Our hot dancer chicks slept-in and we had to hire highschool wannabes instead
Cause 2: Our 10 year old songwriter's most influential performer is Cookie Monster.

Action: Current controls deemed adequate, no further action required at this time.

How to bum a free trip: The Toyota way

2009-12-11T18:21:00.000-08:00

The pinnacle of success for an automotive engineer is to be nominated "chief vehicle engineer." This designation means that an entire organisation and its shareholders place their trust in you for delivering the optimal balance between cost, performance and styling. If you're human, you have good reason to poo in your pants. From now onwards, you will need Winston Churchill's sleep depravation skills minus a mid day nap. Your wife will probably leave you and the likelihood of developing a stomach ulcer will increase by a factor of 450.

So how does one compensate oneself before this arduous task? As always, turn to the Toyota Way. When Yuji Yokoya was nominated the chief engineer for the Toyota Sienna, instead of stocking up on Alka Seltzer, he did the wise thing. He went to his boss and said "Yo Hiroshi, you know that genchi genbutsu deal you're always going on about? I got it. I'm gonna drive the Sienna all over North America. Every f*cking state. Then I'll really understand what people want out of a minivan."

So Yuji did just that and boy did he learn:

While eating hotdogs at Varsity in Georgia he learned the meaning of the pant size 5XL. So he ordered the seating suppliers to make bigger seats. While spying on the competition on the intersection of Rotunda and Greenfield, he figured out that Michigan roads require extra durable suspensions. On route 66 in Mississippi (America's capital of fat) he figured out that Americans are more concerned about their vehicles' number of cup-holders than the number of cylinders.

53,000 miles and 69 tera bytes of pictures later, a well tanned Yuji showed up on Woodridge drive. He just had a company sponsored dream holiday. Now he was ready to work. No wonder Toyota make such great cars.

Old men drive pick-up trucks

2009-12-08T18:28:00.000-08:00

Remember the days when Buick cornered the geriatric market? Well, they lost much of that market share. As Buick drivers kicked the bucket, the oldies filling those gaps came in with a new fashion: the pick-up truck. Commuter pick-up trucks are stupendous tools to drive your wife to get her hair permed or to take a sack of leaves to the leaf-depot. They're also wonderful to reminisce about your younger days when you'd haul 3 chords of logs to burn all winter in your woodstove. Deep inside you know that the retirement home takes care of all that for you, but you never know, do ya? Who knows when the next black out will happen?

So here's the idea of the week: a multistory truck running board with large ergonomic up-down buttons. Compatible with most major brands of walkers.

F.M.E.Atheists

2009-12-03T16:30:00.000-08:00

It's not often that one hears the atheist's voice. They don't believe in stuff and they don't even care to discuss it. Now and again, when there are traces of religion seeping into state institutions, they take a break from pre-marital sex and go on recruitment drives and publish stuff on city buses like "Atheism - sleep in on Sunday morning." Then they go back to eating medium rare steaks on Fridays. However, on the whole, their guiding principles are sound and remarkably well aligned with regular people.

The same goes for automotive design engineers. Despite being obsessed with preventing things from going wrong, most of us don't believe in FMEAs as having any practical value. However as polite and professional individuals, we coexist with the TS16949 enforcers. As long as they don't read past the second page of our FMEAs, we tolerate them. Unfortunately, they often misinterpret this as recognition of how important they are and proof of what a valuable tool FMEAs are. As a result they try to help us out with added FMEA functionality. "Now let's see Pawl, how do you know that the PFMEA captured the DFMEA failure mode of connector interfering with carry over bracket? Ahhh brilliant >add geek laugh here< we need you to create a coding system to check that every line in your DFMEA matches your PFMEA." You'd love to punch your TS auditor, but you can't because he holds the key to your business.

As more managers started seeing FMEAs for what they are worth, the Quality ninjas grew concerned and turned to the NRA for inspiration. They took "Guns don't kill people, people kill people" and coined their own version:

"The problem is not the FMEA. It is the engineers who misuse them"

Obviously if automotive suppliers hired more engineers, we could have some of the prettiest FMEAs in the world. But engineers' salaries are pretty hefty. Good or bad, the industry can only remain competitive if it it minimises the number of engineers needed. The less engineers involved in the development, the more likely it is for design issues to be missed. FMEAs rob engineers of a very precious resource: Time. Time to study designs, time to review designs with peers and time to test. I don’t need a paper to tell me how to prioritise my issues. I just gauge by the number of hours of lost sleep. I don’t need the FMEA process to tell me that I have an issue: I already know what the problems are. My only problem is that I don’t have time to fix them because I have a powerful problem prevention tool rammed down my throat.

Q:Why do volts have heated rear windows?

2009-11-26T17:41:00.000-08:00

The other day I was mowing the lawn with my two year old rechargeable battery lawnmower. I had 10 square feet left to mow when the motor went too limp to have any cutting power left to chop through some seriously healthy dandelion/clover combo.

So I pushed it into the shed, plugged it in and proceeded to go indoors when I saw a strange sight on the sidewalk. An old handicapped lady was pushing an electric three-wheeled buggy along the sidewalk. So I stared at her. When she stared back I asked her what she was doing. She condescendingly replied "I'm pushin' my buggy back home. Battery crapped out on me." That instance a bulb went off in my head. "old lady" I said, "are you real or are you an apparition of things to come once electric vehicles become mainstream?"

Therefore, this week, like a good engineer, I'm sketching some improvements for the Chevy Volt:

I'd love to hear ideas for improvements to this design. Or perhaps you're a spoil sport quality engineer and you could tell us why it won't work.

PS I love the volt. It looks great. However, I find it ironic that it is being made by the company that teamed up with Philip Morris (booo for the baddies) to create fake grass roots organisations that coined the word 'junk science.' Their first use of the term was applied to cast a shadow of doubt on the science of 2nd hand smoke causing cancer. Next was the science behind CO2 causing global warming. Both good reasons to pass on any GM car.

PS2 OK OK So my idea is junk because to use GM's words, Volt is an electric car that can create its own electricity. Plug it in, let it charge overnight, and it’s ready to run on a pure electric charge for up to 40 miles(3) — gas and emissions free. After that, Volt keeps going, even if you can’t plug it in. Volt uses a range-extending gas generator that produces enough energy to power it for hundreds of miles on a single tank of gas. Hopefully, Fritz Henderson will end up at Black and Decker and apply the Volt technology to their lawnmowers :)

PS3 Re "gas and emissions free." Electric vehicles are great news for the environment and open up many avenues towards cleaner energy, but what the hell do GM think is coming out of the chimneys below. Flatulance?

Structural ribs for added pleasure

2009-11-19T18:31:00.000-08:00

You have been working on this electronic GFX module design for 4 months now and your customer painted you into a corner: The carry-over cooling duct is running right where your attachment ought to be and you wish you could gain another 5mm. There is no money to retool the duct. Neither is there money to prototype the module and see if the FKU405 chip will really overheat due to the reduced heatsink size. So you do what you've learned to do best. Make do. You call up your vendor and order the 15 FKUs that you need. The engineering sample limit is 10 but you lie about the projected sales volumes and tell them that it's for a cross platform module with forcasts of 1.25M/annum. The next day they hand deliver a reel of 50. You build a sealed enclosure for the FKUs and hand them off to the laboratory.

You give George the lab tech the instructions: "I need you to run 1000 power thermal shock cycles on this prototype."
"Can't. There's 17 XD547s in there. The sliding chamber won't be free for another 15 days"
"George, just shove this module in with them and make some space"

A few weeks ago I wrote an email to George's boss praising him for his excellent work. As gratitude George gives my tests slight preferential treatment.

The next stage is the design review. With all levels of the organisation out to ensure that their own interests are protected, you know that this will be a tough hurdle:
Purchasing will ensure that all your components can be dual sourced. Operations will make sure that even monkeys can put this together. Your boss, the design manager, will want to challange a few points for the sake of contributing to the review.

Now this is where many years of design experience come in. This is probably the single most valueble lesson that you will ever learn about mechanical engineering:
Never ever put ribs and radii on parts before a design review meeting.

As soon as everybody sees your 3D CAD design projected on the wall without ribs they will go into convulsions. The molding people will say "give my radii or the part will warp!" The assembly guys will say "give me radii or the snaps will break!" Finally your boss will say "I think you should add ribs for more structure."

You see ribs, like some of the rubber Johnnie ads claim, give added pleasure. A pleasure that counterbalances the sense of uselessness that bosses are prone to feeling: "But now I'm at ease, I recommended ribs."

In the mean time you created a great distraction for the fact that at 16V you exceed the maximum power rating of the FKU and that the screw head holding down the PCB is only 2mm from a surface mount capacitor. You know that this will come back to haunt you, but at least tonight you can go home at 5pm and be with your kids.

Giveback café

2009-11-17T20:02:00.000-08:00

To the majority of people, giveback is what rock stars do to thank their fans for making them millionaires: They take some of their wealth and use it to build a school for girls in Africa. In other words, they are giving back to society. Or perhaps on our level, your alma mater requests a donation and you give-back $20 with a note how you`d gladly give a lecture on 'why not to become a quality engineer' to final year students. All good charitable stuff. Another way of saying thank you.

But no, this is not the giveback I rant about today. Today I write about the dark givebacks.

You received an email with a quote package: a work-in-progress CAD model, a statement of work that refers to parts that are not in the CAD model and a request for Quoting that is due in 5 days time. In those 5 days you need to sit down with the design team to figure out what`s needed, get them to write some description for the procurement department which Sally can then use to get quotes. You will also sit down with the manufacturing team to see what new equipment and fixtures are required for this product. The details are sketchy and you`re nervous about the whole quote. Your instinct kicks in and to protect yourself, you start adding safety factors galore. But you also sense that some components like the MB99440H are overkill and would quickly be cost-saved out. So this time your instinct kicks in the other direction and to ensure that you would keep most of the cost savings for yourself, you underquote this line item.

After 5 days of hard work the quote is done and you present it to the president for signing. He knocks off 12% as we really want this business and the numbers look a little bit high.

Two weeks later we`re awarded a 5.8 million dollar contract. It helped that the incumbant supplier is as bankrupt as the customer.

Two years later we`ve been producing the module for 3 months and although we had some pretty embarrasing launch issues, things are looking OK.

But then you get the letter: "...as a result of this situation you have the option to either reduce the cost of your part by $2.20 or give us a onetime payment of $1.6 million dollars."

Essentially this is extortion where despite having a contact, your customer demands a giveback or else they will resource the part, software and tooling to another supplier. It's easy for them to do because:
1)they read your financial statements and they know that unlike them, you made a sh*t-ton of money.
2) They know that you only have 3 customers. Loosing one would be devastating.
What choice do you have?

Welcome to giveback café.

Dear Abby, my son wants to become a quality engineer

2009-11-03T17:44:00.000-08:00

You don`t need to be really smart to teach University students. They are more than capable of teaching themselves. What really sets one lecturer apart from the other is their ability to inspire.

Dr Francis, our total quality management course lecturer was one of these lecturers. He could convey his passion for quality systems with great eloquence. His lecturers ranged from failure prevention to quality assurance and he was very good at tying this to the meaning of quality and what it meant to the end consumer. His more advanced lectures taught us how to use our mathematical skills to solve multivariable problems using Taguchi`s method for design-of-experiments.

So like any young hard-working enthusiastic engineer I decided that I wanted to become a Quality Engineer......

Luckily chance had it that I got an automotive design engineering job before I could self destruct.

So what does a Quality Engineer do all day?.

A quality engineer walks into the office in the morning. His desk is old and has some permanent coffee stains on it. Besides his telephone and a picture of his pimply wife, it is void. His chair is exactly where he left it the day before: perfectly centered in the middle of his desk. He has a marker on his chair to help him line it up perfectly. He learned that trick in 5S training and feels superior to everyone else because he's the only guy who does that.

He checks his voice mail: 5 unread messages. 3 are calls from an angry customer asking for progress reports on the rattle issue and demanding an updated control plan for the change that was made to the supplier's screw coating. The other calls are from his boss, the quality assurance manager introducing him to 2 new issues discovered at the XD57 assembly plant. They need immediate action. The call ends with strict instructions not to tell the customer of the recently resourced LCC supplier.

Joe walks over to the coffee machine and fills his mug. The bottom of the mug has years of coffee scum deposits and he recently chipped the rim. A few runs of sand paper smoothed it sufficiently that it no longer bothers him. Joe returns to his desk, takes a sip of coffee and calls up the capacitor vendor. "Hi it's Joe again....listen, we found 3 loose capacitors in our GX module, we need you down here this morning to sort your stuff....No tomorrow's not OK...Shall we get a 3rd party inspection in?"

3rd party inspection is to problem suppliers what the bogey man is to naughty little boys. Let's say, for example that you make parts in Dallas, TX but you supply to Warren, MI. It's often too far for you to react to a problem as fast as your customer would like you to. So your customer engages a local 3rd part inspection agency that sends someone to your plant to inspect all the supplier's parts to fish out the problem ones. Despite charging exorbitant costs, 3rd part inspection agencies seem to only be capable of employing bottom of the barrel workforce who are as likely to damage parts in inspecting them as they are in detecting bad parts.

With that out of the way, Joe walks across the parking lot to the design office in building 2. He doesn't know why, but he likes FMEA meetings. All the literature he ever read stresses the importance of a good FMEA on a successful product launch and Joe was going to scrutinise every single word. The 2 hour review only managed to cover 3 pages out of the 80 page document but there was another meeting scheduled for next week.

Joe then hurried back across to his own office for a meeting with his boss.
"We are into our third week with this squeak and rattle issue, where are we?"
Joe pulls out an extensive test report that Pawl, the design engineer, was compiling. After subjecting about half of the 50 modules to the 5 parameters Pawl believed could have caused the issue, he accidentally realised that one of the modules had a loose capacitor. Excited, Pawl thought "Yippie, we found the source of the rattle, we can stop the tests".

Well not quite. Joe's boss barked:
"we can't tell the customer that this is the problem. They'll find out about the switch to the new capacitor vendor...Pawl needs to come up with something better... something that will divert their attention... tell Pawl to get his ass in the lab and finish those tests"
So Joe, reminding himself that his job is to oversee quality as opposed to make quality happen, phones Pawl to relay the orders. "Pawl, we need something better for tomorrow's conference call. We can't tell them the problem. we need you to finish the tests...Ohh and don't forget to update the DFMEA"
Pawl, rolls his eyes. The next few hours will be a demoralising waste of time and for the 3rd time that week, he calls home and says "sorry hun, I'll be late again...I've got a quality issue to deal with."
Pawl was planning to spend time reviewing a new design one last time before design freeze but there's no time for that. There's quality work to be done.

The effing FMEA

2009-10-13T19:25:00.000-07:00

As an automotive engineer you appreciate how pressure rises steadily as the design freeze nears. Then one week to go and all hell breaks loose: the design studio changes the A-surface, your designer's daughter developed diphtheria and your connector is interfering with a carry-over-from-X465 bracket.

You need help, but your manager is tied up in a 2 day productivity improvement seminar. Suddenly, the quality manager drops the F-bomb: "Pawl, I need that FMEA on my desk tomorrow."

In his mind nothing reflects a good quality design better than a pretty FMEA. He won't sign off on the design until the first page has been scrutinized and the document feels sufficiently thick. Then like a good engineer you pick up the phone and call you wife and say "sorry hon, I'll be late again."

It's 6pm and you still have to run the study that James, your designer with-perpetually-sick-kids was supposed to run but you'll cut some corners because the FMEA demands attention.

So you open excel, pull up that blank FMEA template and start filling up the cell with 'core team members.' You laugh because at 8pm the 'core team' is at home watching TV. However, they'll gladly lend a hand tomorrow morning in questioning whether you have enough prevention controls for upside-down installation of the connector. Late night work makes you irritable and you prepare a cocky answer to the question that you guess the team might ask. "Errrm, there's no problem poke-yoking the connector as you can't even get the damn thing on. Now leave me alone and let me get back to working."

As the best design engineers know, the next 4 hours that you and your luke warm cup of 75p coffee will spend writing a DFMEA are a futile waste of time. It will do nothing to prevent failure modes but it is guaranteed to make you grumpy.

So why are you lumped with it?

Here's a brief history of the FMEA:In the era of space exploration, when science was still fashionable, a bunch of scientists realized that it's pretty hard to send something into space without actually testing parts out in space. This was fast paced, difficult stuff called rocket science and scientists didn't even have powerpoint and Catia. So to make things easier they created a process to prioritize work based on the severity of an issue, likelihood of occurrence and ease of detection. By having experts in a variety of fields review these documents, they could ensure that obvious failure modes and their effects were accounted for, hence FMEA: Failure Mode and Effect Analysis. This way, if a design or manufacturing issue caused a rocket to exploded and tarnish America's pride, that would take priority over an issue that would say cause the rocket's toilet to block if too much toilet paper was used.

Good idea right? So let's carry it over to the auto industry.Well things were never quite that rosy in the aerospace industry. After a leaky O-ring on the Challenger let to its subsequent loss, the Rogers Commission blamed "serious flaw in the decision making process leading up to the launch of flight" as a contributing factor to the cause of the accident. One of the recommendations involved promoting the FMEA from a design tool to a system of triggering the upper echelons of the space shuttle launch team into signing off on high risk items. Independent contractors were asked to run parallel studies and for good measure a FMEA was needed on every component even when used in multiple locations.

More good ideas right?

Well not really. What we started off with was a tool for engineers and what we ended up with is a political football that is so cumbersome that secretaries and FMEA administrators are required to help manage this elephant. However since secretaries and administrators have not been faced with decades of emails entitled "your test failed," they often get hung up on the wrong things. In fact, they often dedicate their lives to hunting down engineers and twisting their arms until they come up with proof that the likelihood of occurrence of a "poor electrical connection" is a 5 not a 7. Also, FMEAs have gradually contaminated the 'design review meeting.' Such reviews traditionally involve showing a work-in-progress CAD design or rapid prototype to a group of experienced engineers and managers who then constructively point out flaws in a design and perhaps even suggest solutions to the problems. However, when meetings started to involve the completion of an FMEA, that synergy of creativity evolved into a tedious paper driven meeting. As a result the only people who seem to show up for these meetings are pedants, bores and quality engineers. Add these 3 together and an FMEA becomes a F***ing Meeting Every Afternoon.

Ok, so FMEAs are useless, so what's the alternative? The alternative to the FMEA document is to do it the Japanese way:

1) Hire lots of engineers when they are young and grow them with the company. After many vehicle generations and countless engineering changes they will have a strong intuition of what works and what does not.
2) Some problems are next to impossible to anticipate. Even the smartest, most experienced engineers get caught with their pants down. If you want to prevent failure modes from occuring, ditch the FMEA and replace it with design reviews and lots of testing.

Thoughts of the day:Automotive financial success is inversely proportional to the importance placed on FMEAs.

The first 5% of a FMEA occupies 95% of the review time.

The FMEA is the best gift the engineering world has ever given to prosecuting lawyers.

How I Learned to Stop Worrying and Love the Layoffs

2009-10-13T19:19:00.000-07:00

In this day and age, working in the automotive industry is like a game of survivor... without the million bucks. But there's 2 silver linings to all this (at least, to those of us still employed):

1) A big company is like a big fish. It eats lots of food, some is good and some is contaminated with heavy metals. The good food comes and goes but the heavy metals stay, so that eventually it rots the meat. Employees are like fish food. There's lots of good ones coming in, but many of them leave for greener pastures once they discover that high pressure work in an office with 1980s hand-me-down furniture is not a dream job. But the lazy bums will cling on. They know that if they put off tasks long enough, people will stop asking them to do stuff. In fact, their peers will come to realise that they're better off without them. Yes, every office has its 'Wally's. After a few years of this, they are guaranteed easy exit from the office by 4:30pm and no one really cares if they take 2 hour lunch breaks or write 10 chitchat emails a day to their honies in the quality office.
So when layoff V1 comes along, these guys get the boot and productivity actually goes up a lot. After the initial bitterness that comes with any layoff, you're left with a sweet aftertaste. Unfortunately once your company starts hitting layoff 3.0 and there's no slackers left in your garbage pail, that bitterness lingers on for a while.
2) Perhaps these bad times eventually caught up with you too. Initially, when you were temporarily laidoff or compelled to take a pay cut, you were probably mad. To retaliate, you probably said "screw this FMEA... it's 4pm and I'm heddin' home to a nice cold beer." But what you discover is that you're happy. You have no reason to be guilty that you're at home on time. No guilt for setting aside overdue tasks in place of sitting on your deck on a weekday with a beer in your hand. Screw the pay, this is the life.