Hire a Generalist

Back in the day, and I mean a long way back, professional people weren't considered entirely competent unless they were well versed in many fields. I was recently introduced to someone who flies airplanes, builds musical instruments, writes poetry, and runs a tattoo parlor. The person was referred to as a "renaissance man". We think of the diversely-educated and talented as being out of their time now, in a day of specialization.

The generalist has a different way of looking at problems. The specialist, by nature, looks at a problem in terms of his/her specialty. If the problem cannot be seen in terms of their specific approach, the problem is difficult. If it CAN be seen, in any fashion, in terms of their speciality, the solution is shoehorned in, and other possible solutions ignored. The generalist has a more global approach, and looks at problem-solving from a wide open, anything-goes position. There's a wealth of potential solutions to any situation, and only by being open to the unusual solution is one able to sort and apply the BEST solution.

It's difficult to find an engineer these days who is not specialized and whose work experience spans more than one industry. Someone who can speak mechanical, electrical, and materials, the macro and the micro, is rare indeed. These folks are worth seeking out, however, if true innovation and breaking new ground are the goals of your product idea.

Book: Universal Principles of Design

This is a great resource for the student of Design. The book, by Lidwell, Holden, and Butler, is a simple premise, devoting just two pages to each of 100 basic design principles. But within these introductory descriptions are footnotes leading the interested reader to a wealth of research, history, and further reading.

The 100 principles range from the simple and obvious, such as Alignment, Symmetry, the Rule of Thirds, etc., to more obscure and limited-use principles such as the Fibonacci Sequence, Orientation Sensitivity, and Common Fate. Many of these principles are described in their psychological effects to give a basis for their importance, and it is left to the reader to delve deeper and understand implementation in design.

This is an excellent overview of how designers think and what considerations lead to products and their usability. Highly recommended.

Pushing the tools.

I suppose running a company called Strangineer Design is asking for trouble, to some degree. The name and target market were chosen carefully, and as I said in earlier posts, it is working as planned. I get very unusual projects to work on, and a lot of these benefit from strange engineering approaches. What isn't strange is the design and modeling software I'm using.

SolidWorks is a great program, and to design an injection molded part or a machined aluminum bracket, the tool is so optimized that it's become a little too easy. Where the tool is specific to an industry facet, such as for weldments, sheet metal, or molded components, thousands of development hours and input from countless users have made the software pretty much ideal with all the bells and whistles you could need.

But what about a stuffed animal? What about a sheet of fabric as a structural component? What about an assembly whose supporting framework is a folded cardboard box? These are just some of the strange engineering approaches that make for great innovative products, but become obstacles when it comes time to model in 3D CAD. Bringing a lot of experience to bear, and years of testing the limits of software, there are ways to coax these SolidWorks modules into creating objects they weren't designed to create. Even compromises like not being able to assign the mass properties of fleece can be worked around with good ol' ingenuity and hands-on conceptualizing.

Just part of the job everyday here at Strangineer Design.

Presentations, part IV

Obviously, I could go on and on about the subject, but a few last thoughts before I leave the topic for now.

Using CAD software to model parts and assemblies is powerful, and driving tooling directly from the CAD files, generating and tracking changes throughout a product line, all the manufacturing control that CAD has allowed is great at the end of the day. But for the early part of the day, when ideas are flowing and a concept is aborning, think about CAD modeling as a prototyping and sketch tool as well.

There's no reason to wait until the idea is fully formed and then make plans to turn it over to a CAD guy as a Big Step in the process. If you find a contract designer/engineer who you can talk to, who you feel is genuinely interested in being part of the development for you, start modeling. The images of your geometry with symmetry, the right size holes, and perfectly rounded edges will do wonders for making the ideas real and generating the next level detail in your mind. When you start seeing how the parts really fit together, it's almost always a big step onto the next plateau.

If you're an inventor or start-up, remember that corporations throw whole design teams together to get the best ideas on paper and on CAD screens for collaboration. Expand your world by getting a good designer on board early, there's no substitute for an experienced set of eyes coming in fresh and adding to the excitement.

Presentations, part III

Modeling a new invention doesn't have to be a major milestone and expensive step in your product development. It may make sense to do some limited modeling as your invention develops, without investing a lot of money. By taking some portion that is fixed or likely to always be a certain configuration, and modeling it early as a presentation, the rest of the idea can be better visualized and build on that completed foundation.

Having a basis for building up the options and further ideas makes a lot of sense if you are a visual person. Having the standard front, side, and top views to sketch on can be valuable, as well as various isometric or rotated views of challenging areas. The modeled geometry does not need to be sketched over and over as the ideas are evaluated. The structural skeleton of a device may be the most detailed portion, and a model of that structure a great early tool to keep the ideas flowing.

Another great presentation tool is eDrawings. SolidWorks makes this program available as a free download, and the program lets you open native SolidWorks files and spin them in 3D, take measurements, and place comments for review.

If you have some of the invention done in your mind, consider getting that modeled, it could be a great inspiration to completing the work, and the money spent isn't wasted as changes to models are quick and cheap once they're in the computer.

Presentations, part II

Presentations can be low-cost affairs, if manufacturing and fastening details are omitted, but that doesn't mean they can't become complicated. Often, a demonstration of the many options available within an invention can make the modeling a challenge.

If many components need to be swappable, or many configurations of components need to be shown, the models can become time-consuming to create and assembling the visual representations a bit tricky.

Fortunately, SolidWorks is a great tool for creating multiple configurations, both within a part and for overall assemblies. With proper planning, a large number of options can be demonstrated and a presentation created that shows off the ease of changing those options. With exploded views, animation, and the use of color, complex presentations can come across very clearly to the intended market, which is oftentimes marketing and manufacturing folks with very little time to spare for a long-winded explanation of the new device.

Planning is key to putting together a product presentation model, and a lot of discussion up front can make for both an easy target to achieve and a powerful tool for communication.

Levels of Presentation

Many ideas can be represented without fully modeling manufacturable parts. For a more cost-effective approach, leaving out the fasteners and mounting schemes, or the internal parts or features, can get you a "presentation model" showing just the key geometry and interface, for example. To engage a manufacturing partner and get their input and buy-in, leaving out such details gives them something to contribute and latch onto as a partnership develops.

Another reason may be for prototyping. A lower level of detail may make a prototype cost-effective as well, as long as the lack of features does not interfere with the functional aspects of interest. Testing only the critical part first may make for a lower initial investment in both the modeling and the prototyping.

The early stages of realizing a new invention are full of options and choices, many driven by the wary investor. Talking through the options and having cost-effective steps laid out is important to getting off on the right foot.

Designing for the Inventor's commercial success.

I do a lot of projects for inventors who are interested in developing an idea, typically as a sideline to their other profession. Their intentions are to develop professional presentations and drawings to sell the product idea to a manufacturer of similar goods. Coming from other industries or trades where they're not immersed in commercial manufacturing, one of my jobs is to interpret the inventor's idea and tune it to the target market, and so present the design in terms that fit the target company's processes and sales channels.

The trick is to not be too specific in the interpretation. If the idea is a good one and stands on its own, little embellishment or tuning is required. A good idea makes its own case, is easily seen in various configurations and its fit and functions are obvious. For some ideas, making it look a part of the existing product line is essential to communicating how this good idea can transition to saleable line expansion.

Leaving room for the manufacturer to put their personal stamp on it is important. The company needs to tweak any design for the materials they stock, the processes they employ, and most importantly to take ownership of the product as it becomes part of their sales and marketing focus. Design is a subtle thing, and these factors need to be considered.

More on SolidWorks 2008

I have to admire the SolidWorks team for their development process. This release represents a major change in the user interface, the very thing that made the software such a force to be reckoned with in its early days. Back when Pro/E was king, adopting a Windows look and all those toolbar buttons was the way to go, and allowed SolidWorks to take over the MCAD world in a walk.

They've now seen the light as the direction of operating systems and computer users and applications has evolved, and are embracing it while making great advances in de-cluttering. The command pop-ups are smart, making good use of the decade-plus of user data to get the commands that are useful and necessary to appear just as they're needed in the work flow.

Is it revolutionary? No, it's a big change for sure, and one that will put off a lot of users for weeks as they sort it out. But it's the natural evolution of a tool that wants to stay the leader in design and engineering software.

SolidWorks 2008

My local reseller, MCAD Technologies, put on a preview of the new release features coming with SolidWorks 2008 last night. It was a lovely get-together and full house, and the presentation was met with many oohs and ahhs of delighted and happy engineers.

Since I recently blogged about the sad state of drawings, I feel obligated to point out one new drawing tool demonstrated that just about knocked me out of my chair. It's one of the new entries in the SWIFT technologies that were introduced last year, and it's called DimXpert. Essentially, it allows you to pick datums and features on your completed model for the creation of manufacturing dimensions, which are created independently of the feature creation dimensions. This is powerful, for it let's the engineer design the part geometry using the best modeling tools for the job, and then translates the 3D creation into just the correct dimension scheme for manufacturing the part. Those DimXpert generated dimensions are the perfect intermediary step before populating a drawing sheet with views and the cleanup of those dimensions becomes a small cosmetic matter.

I am of course neglecting scores of other features and tools that were demonstrated, many to do with basic modeling improvements, and many specific to modules within SolidWorks. It looks to be quite a major release, everything from interface to functionality having been improved. We'll see in the coming weeks.

Book: The Design of Everyday Things

Originally published as The Psychology of Everyday Things in 1988, they never should have changed the title. This book by Donald A. Norman is not only quite dated, it has little to do with design. I take that back, there are about six very sound and valuable design principles listed very early in the book. The author promises to expand upon them. In fact, he does little more than list them again, repeatedly, amid his curmudgeonly rants about bad designs he's encountered in the world.

The book is more about the psychology of the worst sort of customer, so from that perspective it is actually an invaluable aid to the construction of what is called in the Medical Device industry the Hazard Analysis. This document is an essential part of the product definition and design process, in that, if properly constructed, it identifies all possible errors and misuse, and allows for their consideration in the final design.

A good Hazard Analysis, no matter the device or industry, is important to author and use in the development process. If you can understand the range of possible user, and understand their possible misuse, frustration, fear of failing, etc., the end product can address these before they manifest. I believe Norman is that ideal frustrated user, and his attempts at pointing out design flaws does us a different and better service, pointing out the customer we all design for and want to please.

For that, and that only, I recommend the book. Don't expect to learn much about the design of everyday things, though.

How to get to 2D from 3D.

I can go on and on about the sad state of Engineering Drawings today. Having the module built into your 3D modeling software does not make one a draftsman, and all your engineering skill and experience don't count unless you've specifically studied the art and have the skills to produce good drawings.

If you've done that, you know what portions of the module you can take advantage of, and what of its tools make the task easy to complete.

First of all, start with a clean layout sheet with title block. Whether it's the client's or your own, a good title block is worth spending time setting up properly.

Second, eschew the automatic placement of views. Anyone with experience knows what views matter and it's just as easy to drag and drop. This gives you the opportunity to look at the features first, decide how they need to be dimensioned, and choose the views and scale wisely.

Third, and probably most important, don't let the software populate the views with your dimensions. Some may argue this, but I'd rather carefully place all the dimensions that are needed, rather than start erasing and dragging the piled-up passel of dims and lines and arrows that the software usually provides.

As you place dimensions, the detail views, section views, etc. that are necessary make themselves known, and can be added on the fly. Take advantage of the ability to move and scale views at this point, and the drawing can be made correctly - legible and completely defining the part for anyone who needs to fabricate from it.

What's happened to Drawings?

I love good drawings. The 2D drawing - showing front, top, and side, plus details, cutaways, and all the dimensions needed to manufacture the part - is an art form, and one that is dying.

When it was done on a drafting table with pencil and eraser, highly skilled practitioners could envision the final layout, choose the right size drawing paper and start laying out views. Once finished, you'd be amazed at the placement of detail views, notes, and dimensions, with just the right white space and legibility.

Then AutoCAD came along.

Suddenly, Engineers could do their own drawings with the software. It didn't take any skill or forethought, views could be picked up and moved, scaled if you ran out of room, and the quality of drawings started a long slow decline.

Then 3D modeling came along.

The final blow to drawing quality is the drawing module of 3D modeling packages. Now these modules will create the standard three views for you, and drag along all the dimensions created in the model, boom, there they are on the drawing. Except they aren't right. Unless the Engineer creates the model with good drawing practices in mind, which most don't, there will be redundant dimensions, too many or too few decimal places shown, dimensions shown on the wrong views, essentially creating a useless drawing.

I love to send drawings to suppliers and have them complement me on the quality. The vendors and subcontractors and fabricators of the world have seen their job become harder as they receive drawings they have to interpret. I believe the demise of the drawing has driven the adaptation of tooling that reads 3D model data directly. What's the alternative, sifting through all these terrible drawings to make heads or tails of what's intended?

When a supplier does need a 2D drawing, and they take time to call me just to say "thank you" for a real quality drawing, like they used to see back in the day, then I am happy.

Style in a 3D CAD world.

SolidWorks is a great program, make no mistake, but how does such precise and perfect geometry fit into a design world looking for style? Aren't perfectly straight lines and curves a constraint? It does pose some interesting challenges. Of course, if the style or design is calling for geometric precision, that's easy. A pattern of repeating angles, circular cutouts, scaled ellipses, etc., are all child's play with CAD tools.

The challenge comes when the geometry needs some funk. A perfect sphere is easy, but make it look like it's melting a little on one side. Or blend the rectangular beam into a heart-shaped balloon. Fortunately, SolidWorks has the tools for these kind of artistic flairs. It takes some creative application of some of them, but there are ways to do just about anything wonky, and there's no reason to blame the 3D CAD environment for any limitation. Add-on programs abound as well, as people in a wide variety of industries, as well as artists, have found work-arounds for their needs.

So, if it doesn't look like a CAD project, that's okay. It can still be, and turn out great.

The Lowball Game

It is said, in business circles, that the worst thing a new business can do is get caught up in the price game. You can never win. Especially in this business, there is always some kid who has SolidWorks on his home computer and is willing to do a model for next to nothing. Maybe for a school project, maybe just for fun, maybe for gas money.

I get inquiries from all manner of inventor and part-time tinkerer and one of the first things I have to do, though it's not always pleasant, is find out if the client is serious and has any idea what costs they're looking at for design time, prototyping, marketing, tooling, and manufacturing startup. Without them doing their homework, I'm likely to get caught up in the lowball game, because they're just looking for someone to do the CAD work, and that can be had cheap.

The CAD work I do isn't cheap. It includes setting up a file and documentation so that the project can be transferred or recreated later. It includes archiving so that the data is never lost. And it includes foresight into the rest of the process, something that only comes of having done it for decades.

As a client, don't fall into the trap. The money spent in the front is minimal, even with someone like me. Make it a good investment and don't go for the lowball quote. Is that a sales pitch or what?