Back in my first blog, I reminisced about my days as a draftsman. For this topic I am going to do the same. Back then we used lead pencils on paper or vellum. When we needed to share drawings with the shop, we made blueprints of them using a very large ammonia based blue print machine that looked like it would as soon eat you as make a copy of your drawings.

I remember being mentored by the senior engineers. One of the things they use to tell me was I needed to always consider how something I was designing would to be made. To make sure I did this, they would have me spend time in the shop. While I would help where appropriate, my main goal was to observe how things were being manufactured. I would watch everything from machining to assembly. This was invaluable experience, as from that point on I would do my best to always ask myself on new designs or design change, “Can this really be made?” And for the most part, I could answer that question. If I couldn’t I would be sure to get with the appropriate people in manufacturing to find out before calling my design complete.

In my current position I now get the opportunity to see various engineering departments. In many places I see a big disconnect between engineering and manufacturing. There seems to be this big virtual “Wall” between the two departments. The engineering group develops a new design. They usually run through a very formal design process. And with new CAD tools like PTC offers in Creo 2.0, they can run through various analyses, simulations and interference checking all from their desk. This new capability as proven to help reduce the number of pilot runs and rework required due to design flaws. Unfortunately, with all this new automated checking, I think sometimes the engineer loses sight of how their design will actually be manufactured. In many cases I have noted that the designer really has no idea how manufacturing actually gets their designs built.

Too often a design goes through a rigorous design process, only to be “thrown over the wall” to manufacturing. Once there, the manufacturing engineer often would require changes. Best case, if they had a good PLM system like Windchill, they would start a formal change process, asking engineering to make the required changes and send the design back to manufacturing. This still costs time and money, but all CAD models will stay in sync with how the product actually gets built. However, in many companies, I see manufacturing modifying the CAD models or assembly to reflect their needs. Usually they then save their version of the objects on a local drive or network drive. These files are then completely uncontrolled files outside the companies PLM system. However, without them the product could not be manufactured. Just as important these manufacturing files do not match the designed version of the CAD files. I am not talking the “as designed” to “as manufactured” bill of materials. These often are different, and they should be. I am talking the CAD models themselves being different. This potentially is throwing all the analysis and testing done in design out the window.

I am sure for most of you I do not have to explain the risk of having uncontrolled manufacturing version of CAD files. Why don’t more engineering departments and individual designers today take manufacturing more into consideration when designing? I have a couple of opinions on this. One possibility could be the shorter and shorter design timelines engineering has to work with. They just don’t have the time anymore to research their companies manufacturing practices. Once again I challenge management to truly consider the time they supposedly save in engineering, to the risk with how things truly get manufactured. Another very good possibility is more times than not manufacturing does not happen anywhere close to the Engineering group, many times in completely different countries. What I used to be able to do when I started out, is no longer an option for many engineering departments.

I can’t say I have a complete answer to this growing problem in our industry today. However, one thing I have seen work well is a true design review meeting that includes representatives from manufacturing. Before releasing any design, it must go through one of these reviews. Any concerns with how the design will be manufactured can be brought up during this meeting, and addressed prior to design release. Another option I have seen is creating an “As Manufactured” or “As built” version of the CAD files in a PLM system. Sometime the manufacturing engineer makes the changes, sometimes they are sent back to the design group to be made.

Let’s tear down that wall! If you’re in engineering, consult the manufacturing group about your designs. If you’re in manufacturing, let engineering know you need to modify files just to get them built correctly. Talk though it, bring attention to it. If you don’t, no one else will. You may not think it affects you leaving things function as they are. But, if you are not helping you company become the best it can be, in the end it will be the employees that inevitably suffer.

How does your company deal with changes required to a design so it can be manufactured? Does your engineering group consider manufacturing during the creation of their designs? How do you train new engineers/designer’s manufacturing methods? Specifically, if your manufacturing facilities are offsite. Does anyone else feel this is a growing problem in engineering/manufacturing companies today? While I get to see many companies, I obviously do not have visibility into every company in the country. So, please, respond with your thoughts, opinions, and how things are done where you work.

No, I do not mean the game. I mean those engineers “back in the day” that would make sure every drawing released from engineering was as correct and accurate as humanly possible.

Longer ago than I care to mention, I started out in the engineering field as a draftsman. I started out using lead pencils and electric erasers, if that gives you a clue as to how long ago this was. In a very large room with many drafting desks I was the low man of the group. At the top there were Checkers. These were the top of the engineering group at the time. They had the most experience and were some of the highest paid people in Engineering. A drawing of any type could not be released for production or customer consumption without one of the checkers first reviewing it. What stands out the most for me is that there were very few, if any, drawings that did not get some sort of red markup on them. I know, because when I started out I was the guy stuck fixing all the mistakes the checkers found. Keep in mind that they were not only checking the aesthetics of the drawings, but they were also checking all the geometry, equations, mass properties, interferences, and many other aspects of the design. They would even check if it was accomplishing the design intent for the overall project. At the end of the day I rarely heard of manufacturing having to come back to engineering for missed dimensions, or interference issues or anything else for that matter. It did happen of course, but not as much as I see in today’s engineering / manufacturing companies.

Now let’s jump ahead to the current day in most small to mid-size engineering / manufacturing companies, and even some large companies. I can’t tell you the last time I have seen anyone in a Checker role. Yes, I have seen companies do peer reviews, design review meetings and things like that. But there is no one dedicated to insure that nothing leaves engineering until it is thoroughly reviewed for mistakes or issues. Over the last few decades have engineers become perfect? I know of some that may think so, but no…that is not the case. What I see happening is manufacturing is now the group that is indirectly checking the work of the engineers. How many times have you seen manufacturing coming back to engineering to request a dimension change, or with other issues with the drawing or design? Sometimes it is minor issues, but other times it is a major design flaw. By the time manufacturing finds one of these major issues you could have significant money already spent on tooling, raw materials, man hours, etc. Think about how many recalls we have seen from major companies over the last few years? How many mistakes are not caught by manufacturing at all? In many cases it is not until the product is built and shipped before the issue is found. Think about the cost to the company in those cases.

The fact remains that in the overall corporate view, you are now relying on manufacturing to make sure you are not releasing flawed products to the world. As far as the so called time savings…what do you think takes more time? Having the checker verifying all engineering work, and then making a requested drawing or model change before it is released…or…Releasing designs to manufacturing, then manufacturing finds an issue. Now a formal change request is sent to the engineer. A change process is started. Dispositions must be made for any raw materials already acquired and so on and so on until the change has been properly addressed. If someone took the time to analyze the real cost behind all minor and major changes, taking into account man hours, material and time cost, I think you would be surprised at what you would find. Assuming a checker would catch 75% of these issues before they were released, I am betting the amount you would come up with would cover the cost of the checker role.

Some of this checking can now be done with today’s modern CAD software, or add-ons to them. Such as Creo and Model Check. So some of the more basic things checkers used to check can be done with software if configured correctly. This would save what would be the modern checker significant time. Plus there are even more tools available that require human interaction that could help them check and review designs more efficiently. Like Creo View Design Check. So the time it did take checkers could be drastically reduced, further justifying the need for this role again for all sizes of engineering companies.

I am not naïve enough to say a checker would catch all mistakes. Even a checker is human, mistakes will still happen. I am saying the number of engineering mistakes released would be dramatically less.

I would love to hear comments or feedback on this. What does your company do to ensure engineering releases are as accurate as possible? How much is caught doing peer reviews? How thorough are your design reviews? How many minor undocumented changes are brought to engineering by manufacturing to be corrected? How many major issues does your manufacturing department find? Do you feel there is still a need for checkers in today engineering companies of any size?