Design for Assembly: taunting from the manufacturing floor

When it comes to design, sometimes what works in the uninhibited realm of solid modeling (where just about anything is possible) doesn’t make a bit of sense in the real world. A good design engineer knows this from experience. A bad design engineer may be lacking that same experience and may be in need of some hard feedback in order to improve. While you could pore over Powerpoint error count metrics and root cause analyses to systematically highlight specific design deficiencies, there’s a far simpler way: a good old-fashioned hasslin’ from “Bulldog.”



The legend

His given name was William, at least I think it was, but he chose to go by the more colloquial alias of “Bulldog” and his company badge confirmed that it was somehow not a nickname. Bulldog looked like hell, like a grizzled old prospector crossed with a post-apocalyptic warlord. It was easy to imagine his time off involved casually punching bears, long walks through factories full of burning tires, and breaking lesser engineers over his bad knee. He swore profusely and smoked like a cigarette factory. His motto was writ large on an oil-stained mural proudly taped to the back of his toolbox on the hangar floor:


Above all else, however, Bulldog was a damn good aircraft installer, a true artisan. Probably one of the best. And he knew it. Over the years, he had seen just about everything and had managed to profusely swear down several generations of, what he termed, “those $%@# engineers.”

Adventures in first article liaison

One of the earliest jobs in my career involved a facility with onsite aircraft assembly and installation work which benefited greatly from a strict policy: the original design engineer was always responsible for their first article liaison. While this policy ensured design intent was properly communicated downstream, it also had a more lasting benefit: you learned through real feedback when your stuff worked and when your stuff didn’t. And when stuff didn’t work, that meant a half-mile long march down to the hangar floor to respond to squawks. Every one of those marches was an opportunity for someone on the manufacturing floor to give an engineer the what for.

This is how I would meet Bulldog.

Especially hilarious was the fact that I was a green co-op at the time (a non-degreed engineer-in-training on internship between semesters of school) with authority slightly less than that of a mild fungus. He would look at me, amused at the fact that I was likely younger than many of his grandchildren, and quickly expressed his particular disdain for unnecessary complexity. Say, for example, spelling or pronouncing my last name. So along with the help of the rest of the manufacturing floor I was bequeathed with an alternate part number. I was to be Ed Lamborghini. As in

Oh for $%@# sake, here comes that $%@# engineer, Ed Lamborghini.

He would then proceed to dismantle everything he didn’t like about the particular assembly I was working on using a finely crafted barrage of chained expletives for emphasis. And the scary part: he was right. And that meant an opportunity for learning.

Design for Assembly

Bulldog’s well-intentioned hazing and verbal abuse was believe-it-or-not a compact form of Design for Assembly. Design for Assembly accounts for both human and machine factors in assembly, including workflow, manufacturing capability, cost, and opportunities for automation. Design for Assembly is about taking form, fit, and function beyond the context of a single part and into the interaction of multiple parts in a more complex machine.

Simply put, it’s about putting things together more reliably by considering:

  • Accessibility: Can someone actually get in there and do it with the tools they have. You’d be surprised how quickly you can lose a sense of scale when staring at a CAD design for too long without physical reference.
  • Fasteners: Using the right type of fasteners, considering access and clearance.
  • Part Orientation: Understanding how parts are indexed and positioned off other parts and identifying and eliminating opportunities for confusion, misalignment.
  • Simplification: Knowing when it makes the most sense to combine or separate parts for sake of easing or enabling assembly.
  • Tooling: Knowing when and where to cost effectively exert more exact control through tooling aides.

Design for assembly is most effective when there’s a closed feedback loop between the design authority and the manufacturing authority. Something beautifully illustrated by that first article policy and Bulldog’s particular enthusiasm.

An uneasy peace

Over time I would learn to pay special attention to Bulldog’s particular passion for Design for Assembly, and soon enough I would manage to earn his respect. I wouldn’t go so far as to say we were friends, as he’d probably sooner stab me in the eye with one of his larger screwdrivers than begrudgingly admit any sort of minor pleasantry to an engineer, his sworn enemy. But he certainly swore slightly less and stopped hasslin’ me so much, saving his wrath for the next poor engineer.

Such hard no-nonsense lessons are critical to building the engineering experience necessary to properly differentiate good design from bad design. Building such experience is certainly harder these days with manufacturing often outsourced, including artificial separation between design and manufacturing authority. Perhaps Bulldog can come to the rescue.



  • Evan Sullivan

    Very entertaining, Ed. Thanks!

  • Ryan

    I agree design for assembly is harder when tings are outsourced and the assemblers may be in different regions. That’s when the Bulldog becomes the field/service technician and the changes are much more expensive!
    Great article, I enjoy reading your work.
    Oh, and I’d also like to see a phonetic spelling of your last name, too!

    • Ryan,
      Great point, It’s an interesting conundrum. It’s expensive to keep vertically integrated with manufacturing in-house these days, but those field/service changes can carry a high price tag as well. Perhaps the balance is found with more robust prototyping?
      To resolve the mystery, my last name phonetically: lō-pä-‘te-gē If you know your Spanish vowels it’s much easier to guess how it sounds. (This has been a public service announcement)

  • Sanjay Kathuria

    Well written to highlight the significance of Design For Assembly….If DFA Objective is achieved at the expense of Cost Objective, we will still have the lowest Cost Of Ownership as the assembly will have the Best Chance to be achieved with desired reliability…Happy Customer will come back wiith repeat orders as well..

  • Jeff Hapgood

    Ed, Great article. Would you allow me to excerpt some of it for presentations to CAD students and instructors from career techs and high schools? Please let me know.

    • Jeff,
      Certainly! Just make sure to include a citation mentioning myself and GrabCAD.

  • Dave

    Great article and well written. I remember my Bulldogs when I was a co-op and young engineer. I too have learned to appreciate them (after hating them first) and have even become friends with many of them. You did a great job tying the practical experience to the vocabulary the “authorities” like to use.

  • Spacebuffalo

    Ed Lamborghini,

    Can we all safely assume that this happened in the distant past? In modern times (the last 20 years) if any of what you write about happened HR would be summoned, edicts of “hostile work environment” would be declared, and Bulldog would be summarily fired.

    But thanks for the stroll down memory lane.

    • Spacebuffalo (Might need to take that awesome handle as a middle name: Ed Spacebuffalo Lamborghini, nice ring to that).

      Indeed it was quite awhile ago, but slightly less than 20 years, while it was definitely an adversarial work environment I never thought of it as hostile. It’s certainly a fine line. You’re right to say that most modern HR thought would react unilaterally in such a situation. But it’s difficult to artificially simulate trail by fire learning, makes for the basis for an interesting discussion about what direction training should go in the future. I think that Bulldog never intended to cause me actual harm… probably.

      • Spacebuffalo

        I’m sure he never did actually mean you any harm. Too bad we live in a world where everyone is thin skinned. I used to work around a guy who would growl at me every time I walked by him, that is until I mentioned it to him. I guess I had it coming, I wrote “FOAD” on his whiteboard one day.

  • When I got my first job as an engineer I was not allowed to approach a drawing board until I had spent three months in the workshop. Here they taught me how to cut, weld, drill, broach, mill and turn on a lathe. They knew it was insufficient time to learn how to do any of these things WELL, but I would at least generate respect for the guy who had to assemble what I drew

    To this day I still remember gobs of molten metal falling on me so I avoid specifying overhead welding unless absolutely essential. I always prefer two round holes instead of a single broached square hole. I know how much easier it is to tap a through hole than a blind one, etc.

    That was 47 years ago and although I have forgotten most of what I learned at school I still remember clearly what I learned in my three months under the kind tutelage of workmen who so impressed me with their skill.

    • I don’t like the coercive nature of how they did it but I do see the value. Achieving the same end with less of a push tactic would be nice to find but I don’t have the answer.

      • I never felt this was coercive. It was just part of the job. I think the guys in the workshop were even quite amused by it. They never teased me for my incompetence because they didn’t have to make the point. I knew they were all far more skilled than I could ever hope to be and they knew that I knew it.

        Apart from the obvious benefit of learning design for assembly, it also made communication with the workshop more straightforward as we all knew each other by the the time I was ready to send designs their way to be assembled.

        We were making roadbuilding machinery and gravel plants which is perhaps less frenetic than aerospace or automotive.

        In those days the fitters used to visit the drawing office, have a cup of tea and we’d discuss the problem, often involving other more experienced engineers in finding the best solution. It was a more civilised and agreeble way than described by Ed Lopategui, and probably not much less efficient.

      • Did you enjoy that? I’m gonna be perfectly honest, if it doesn’t have aero- or auto- I’m not interested. It just doesn’t intrinsically motivate me. How did you handle getting excited for roadbuilding machinery? Or was it never an issue for you?

      • I find any engineering challenge stimulating.

        When you deal with smaller, less sophisticated projects you have a much greater influence on the success or failure of the product. You can look at it and say “I did that. The machine I conceived, engineered and drew can mill 200 sq.m. of bitumen surface per hour to plus or minus 5mm accuracy”, which is a bit different from saying “Those are my undercarriage doors. Pity the aircraft doesn’t fly so well, but the gear operates perfectly!”

        And as it happens, the company I was working for at the time built the runways at Heathrow and most of the M4 motorway, without which your aero- or auto- bits would find it harder to operate.

        Sorry, I do not want to sound agressive or confrontational. I take your remark in good spirit as I hope you do mine.

        For the last 35 years I have been designing yachts… now they are very stimulating in a totally different way.

      • Yeah, I understand ya.

  • John

    Is this Bulldogs real pic or is it a generic grizzled old prospector crossed with a post-apocalyptic warlord?

    • John,
      The image is generic, any attempts at taking a real photograph of Bulldog by mortal men have all ended in failure. But it’s not to far off. Just a little more prospector and a little more apocalypse.

  • joe

    Shop time should be a REQUIREMENT for a degree in mechanical engineering I have seen tens of thousands of dollars in waste over the years I have seen parts nearly impossible to make. As a shop that both desings and builds parts we try to work with the engineers from our customers when we see something difficult to machine, some will listen some will take an attitude of something like I have 5 degrees what do you have, how about 40 years in machining experience. The engineers that take that “I know best” attitude never last because they design parts that are too expensive to make. It is quite common to see this, fortunatly more of them are willing to listen and in doing so learn exactly why we can not machine certain features the way they are designed. They also learn that every zero behind that decimal point means dollars when they tolerence something. Give a new engineer a drawing of something he just did then put him on a lathe or a Bridgeport and tell him OK now make that and it wont take long for him to learn what can and can not be done. If you have zero experience in the shop you will never make a good engineer or if you do it will take many many extra years of listening to the guys who are trying to make those parts and paying attention to what they tell you.
    Remember this a 1 inch square block of steel with a +/-.06 tolerance would be say $30 bucks give that same block of steel a +/-.0001 tolerance and you have a $300 dollar block.

    • I don’t believe in this, I believe in the power of co-operation and specific people doing specific tasks. Only then can people feel empowered and feel that way quickly enough to become motivating. The key is that everybody work together and that’s what virtual prototyping allows to do so well.

      A problem also tends to be that in generic CAD systems, you design parts and assemblies in a vacuum and it seems to have been a tough struggle to get digital manufacturing established as a tool outside of the big aerospace and automotive companies.

      • joe

        You can not be good at a task unless you know all there is to know about it. Engineers with ZERO shop floor experience know NOTHING about machining in the real world. Without machining experience they design parts that are impossible to make or so expensive that they are not practical. I see this on all the time it is surprising how little they know about something that is such a huge part of what they do. You think a specific person should do a specific task well then get a robot. You want a well rounded person that can take on a project then get an engineer with shop experience because the ones that do not have it quite frankly suck. I have worked with dozens and dozens over the years and sorry but this is just a fact.

      • I liken it to the idea of an assembly line, where people each have specific tasks and these are SMART, you know, simple measurable etc… and a job rotation system keeping things interesting for the people. It’s kind of the idea behind the ’90s Total Quality Management.

        I abolish the idea of the well rounded person that is doing his best, because it’s a utopia that can not be achieved. Well rounded individuals produce waste, specific highly trained specialist and experts working together do not produce nearly as much waste.

        It’s just my view on it, everybody has their own vision on this. Certainly the least that can be said about it: it’s way more complicated than it seems on the surface!

      • joe

        Well I am sorry but you are wrong on that in my experience I have owned my shop for 30 years and the well rounded person does a MUCH better job than the single task person does. The single task person shuts down when presented with anything new the well rounded person figures out what needs to be done and does it . It is not a utopia it is simple reality it is years of experience, the people in my shop that I can give a drawing to and have them make the part using what ever machines that they need to are worth 10 times the person that is a button pusher on a CNC machine and that is why I do not have ANY button pushers in my shop. To put a person in front of a machine to do nothing but the same task day after day is simply wrong that is where a robot should be placed. Human beings need a challange they are not just a task performer they are human beings and should be treated as such. It is not as simple as job rotation sorry but that just does not work you need to challange the mind you need to offer improvement through experience, you need to give people a change to better themselves and you need to reward them for doing so. People that work for you are not just numbers to be manipulated for the best bottom line that is a horrible way to think of them. Without good people working with you you will have nothing it is the people working there that make the company not the owner not the boss. One person can only do so much if you want to grow you need to hire if you hire you want the best and if you can not find that person then you train them and you train them in any and all aspects of the job there are there to do. Anything less is a waste of a persons life.

      • I share your view on treating people as the most important asset within your organisation and the emphasis on continued education and training. However there also need to be explicit processes because if Bob leaves, you don’t want that knowledge to leave the organisation. Every process has a customer and most problems lie with the process not the people, rewarding individuals can have contrary effects because you are rewarding based on the outcome of processes beyond an individual’s own control.

        I’m not a big fan of the shop environment philosophy. But I understand the no-nonsense attitude of a shop environment. It’s just not for me. Everbody is different, I think that’s a key point to consider in every joint human endeavour.

      • joe

        I think you are missing my point if Bob was the only one that knew a certain process that would mean that the other people were not trained all around. If all the people are well trained anyone could pick up Bob’s job after he left. As far as most problems being in the process not sure where that comes from if we a need to use a ceratin process to produce a part we do not use a faulty process then expect the people to work around it . If the process is faulty a well rounded and trained person will see that and fix it where as a less skilled person is more likely to do nothing.

        “rewarding individuals can have contrary effects because you are
        rewarding based on the outcome of processes beyond an individual’s own
        I am not even sure I know what this means it does not really make a whole lot of sense to me.
        I am also not sure what you mean by a shop enviroment philosophy if by that you mean it is up to managment to control everything you are very sadly mistaken a company would fall apart if it was run that way. Managment more times than not is the cause of problems in a manufacturing environment because they are clueless as to what is going on out on the floor. Unless of course they came from that floor, bring in some kid fresh out of school you have a worthless worker until he has earned some experience. It is amusing how smart these people think they are because they went to school for 4 years when that 4 years really amounts to nothing in a lot of cases.
        Is some cases a complete re training is required to make them unlearn the nonsense they have been force fed, it is a real joke in a lot of cases. I would rather hire and train from the ground up than go through what I have had to go through with so called educated graduates.

      • Shop environments tend to have these bromances going on where everyone needs to be thick-skinned and hard knuckled. The element of pride that goes with that is admirable but ultimately I can not agree with that philosophy as a long-term growth engine. Because it works on competition and fear.

        I agree the requirements of industry are ill met by formal education.

        A well rounded person will become increasingly entrenched in habits, do you use model-based definition for your shop? Why is STEP-NC still not an everyday reality but instead a fancy standard few actually use?

        It is the rigidity of so called well rounded individuals.

      • joe

        I am not sure where your experience coomes from but that is certainly NOT the way things work in the shop. I started my company 30 years ago I have worked with hundreds of other companies and I have never seen anything like that it just does not happen. If any of this kind of stuff is happening it is far more likley to be in the office where people have ego’s that wont quit based on what stupid books they have read or how many degrees they have, it is amusing to say the least. The problem is most of those degrees are not worth the paper they are written on.I always have to laugh when some joker comes out with some new method
        of managment or crackpot manufacturing idea that is going to fix the world and how many suckers fall
        for it. “A well rounded person will become increasingly entrenched in habits”?? wow where does that come from, some book, because it is far from reality. It sounds to me like you have very little real world experience in the manufacturing environment and are basing everything on the idea that STEP-NC is worth a nickle. Just because something exists does not mean it is good or it fits every situation. STEP-NC is worthless in most shop enviroments it is just another one of those ideas that someone came up with that is supposed to be a fix all and it is simply is not. This is the problem with having no experience, people think that the computer is the solution to everything and it is NOT and never will be. Technology is great but ONLY if you know where to apply it and STEP-NC has no place in the majority of manufacturing plants. I see most inexperienced people relying on some computer program like it is the answer to every thing, I think they played way too many video games when they were kids. A few years in the shop enviroment and you would understand why this is true, there is no such thing as one solution to fix everything. There are cases where technology can waste huge amounts of time I have seen it many times where people try to force a solution on something that just does not fit but oh it is new so we have to use it, nonsense. Learn to apply the correct technology where it will produce the right results not to use it just because it it there.

      • STEP-NC is love, STEP-NC is life!

      • joe

        Yea OK when you get older and have more experience you will understand. Right now you have not been around the industry long enough to know how things work in the real world. You simply have not been exposed to enough of this industry to have a real understanding of how it works.

      • And how it works is inefficient and wasteful, how could you possibly be against an effort like STEP-NC?!

      • joe

        You do not know what is efficient and what is not you have never run a shop. You also have a reading comprehension problem I never said I was against it I said it only has value when applied in the right way and for the right reasons. You can not understand this because you do not know how a company works. When you get older and have more experience under your belt you will understand. You dont use a backhoe to dig a hole to plant a tomato in, does that help you to understand. I am more aware of current technology than you are, trust me my company dependson it but I also know that you do not impliment something just because it exists that is a horrible waste of resources. There are certainly areas where STEP-NC has value but they are highly limited and in fact in most shops would be very wastefull and not solve a thing.

      • I suppose we agree to disagree, for me STEP-NC is a vision and an all or nothing affair. I suppose we should stop bickering, hehe. Have fun with your job shop Joe, I’m not your governour. I also don’t care :)

  • I think this happens to many new engineers. However, I do not blame them or think they need to go through it. I blame the idiotic organisation structure and culture that leads to designs being done without input from manufacturing.

    I have sensed that aerospace is a bit behind in this area and this story would suggest that I was correct. I think such a story would be mush less likely in automotive where manufacturing and assembly are well known to be make or break for a company’s survival.

    I have however worked in manufacturing companies where there was little encouragement of good concurrent engineering. I actually recall one manager, when I started at such a company, asking me why I was even talking to people in production, quality and service while working on the design. I think he would rather I ask him so that he felt like the expert. That’s when I understood why people in production, quality and service seemed surprised by me talking to them at the early stages; no-one else talked to them because it was not company culture.

    Good thing I was culturally insensitive, and instead applied what I had learned from the textbooks on concurrent engineering.

    • Clint,
      From my experience, automotive has an advantage here because it tends to be more vertically integrated, most of the supply chain are delivering components and not major assemblies. The challenge in aerospace is in partnerships where design authority is distributed across large interrelated assemblies, or in outsourcing environments where maintaining good communications between design and manufacturing functions becomes difficult when they cross export, organizational and geographical boundaries.

      • My experience in Auto seems very different from yours, Ed. I saw huge sharing of authority during the design phase of automobiles.

        I think has more to do with scale and competition. During WWII Henry Ford was able to ramp up production of aircraft using what was standard in the auto industry. If the aero-companies were subject to the same levels of competition as the auto-companies, then maybe that would force them to encourage better concurrent engineering.

        However, this should no be something that requires force. It’s just good engineering.

      • Missed your comment there, Clint!

        Agree that the competitive environment does play a factor in aerospace, particularly on the military and space side of the equation. Old cost-plus mentalities drove the bulk of that, but much of the work has transitioned to firm-fixed, so it should be improving.

        Interested in any specific examples that you are willing to share w/r to shared authority in automotive design. I’m aware of projects like the BMW-Toyota i8, but felt such was more the exception than the rule.

      • All sub-components were predominantly designed by the suppliers – basic input from the automaker.

        Gearboxes, AC compressors, seating, entertainment equipment, cooling components, suspension control mechanisms, that kind of stuff.

        Each supplier shared responsibility with the automaker and they had to work together to ensure manufacturability and leanness. It was basically the norm.

      • Clint, I’m not sure we’re not talking apples to apples. Supplier component design as you describe is quite common in aerospace as well, interface and performance requirements are specified, but internal design is left to the supplier.

        Instead I was attempting to describe when design authority is shared by multiple-risk sharing partners across major assemblies. Imagine if one partner was responsible for the front half of the car (complete with integrated systems), and someone else controlled the back half. Seems strange, but that’s actually how some aerospace manufacturing works. Fuselage sections, wings, cockpit, and empennage may all be built separately by different partners and later joined across common interfaces by the prime contractor.

        Such an approach is likely less practical in automotive, especially with the turnaround for producing new models is considerably shorter (and why platforms are used). Do you know of any examples like this in automotive?

      • Maybe we are and maybe we are not. I would not see the size of the components – half the final product or just a nut – being the major issue. I see that issue as the number and diversity of designers that need to work together is the bigger issue. And the more parts the more important it is that they be designed to fit together from the onset. And I maintain that the auto industry does this well due to the competitive nature of the industry, which demands a congruent culture. If other organisations put more effort into encouraging a culture of concurrent engineering, then we would not need to waste time having conversations about manufacture-ability after the first design with other engineers.

      • I remember reading a blog post about the Chrysler transition from CATIA to NX and in it, a couple old-timers shared their views on the automotive industry that felt very open and truthful. And it seemed to me like it wasn’t nearly as rosy as you make it out!

        In the end I think as long as fundamental structures are not being challenged things will never change. We strive to optimize the products we design yet we seem to largely neglect to optimize the processes that make it happen. I see a lot of value in the old Total Quality Management movement. It seems to emphasize people, in everything it states.

      • I do like the idea of challenging things everyone and then.

  • Designer Dan

    Where I went to school, Lehigh U, we were required to take a 6 week summer course at a local machine shop where we learned to run a manual lathe, a manual mill, a shaper, and I forget what else. We also had to use hand tools to finish the parts we made. That knowledge has come in handy in years since.

    My comeuppance as boy engineer occurred on my first job. I was designing chassis for electronic enclosures, and following the techniques that I had seen my brother-in-law use (he was a EE and an amateur radio operator), I secured the cover plates with sheet metal screws. Got a call from the Manufacturing Manager.

    He used 4 and 5 letter words amply while he explained to me this was government work, and that we used machine threads. He took me out on the floor and showed me PEM nuts, explaining tome that these were welded to the inside of the chassis, and suitable holes drilled in the cover plate to take machine thread screws. I discovered later that PEM nuts and their products were made somewhere down the road from school.
    I came to the conclusion that most engineering instructors were not familiar with the real world after they entered kindergarten.