SolidWorks Certification Process

SolidWorks has had a certification process for a long time. It originated in 1997 with a written test for reseller tech support certification. It had very odd rules. If you got certified, it was actually the reseller organization that got certified, not the guy who passed the test, even though you were the only one who passed. It was obviously to cater to reseller business owners who hired hot-headed rougue engineers who might get certified and then change their minds. Anyway, the whole reseller relationship with SolidWorks is very strained, and fits funny, although it seems to have worked well (business wise) for SolidWorks and resellers. You can decide if it works well for you.

There was also a training certification in 97. That consisted of teaching a sample lesson from the training manual to a group of other test takers and the guy evaluating you.

A couple of years later, maybe 1999, I don’t remember, they added the certification for users, the CSWP  (certified solidworks professional – affectionately called “See Swoop”). I didn’t bother with this until I was working somewhere and they wanted me to proctor tests, so I had to take the exam in 2002.

The format of the test evolved from a paper hand-out for the original tests, to a computer program that randomized questions and answers. Multiple choice and modeling questions.

The test was imperfect, and I was convinced that some questions had no correct answers, and some had multiple correct answers. I thought this was due to people who had a rather narrow and pedantic view of SolidWorks functionality, working only from how the software worked in theory, rather than from how it actually worked. One example was the classic “do inference lines add sketch relations?” The answer they were looking for was “no”, but the correct answer is “yes”. If both ends of a line pick up the same inference to the same entity, it will create a coincident relation. Draw a line from one side of the origin and then make the line go away from the origin. You get the blue dotted line at both ends of the line, and the line is black when you are done – fully defined (except the ends).

Anyway, there were other questions on the old test that I liked because it made sure you knew definitions and names, but these drove other people crazy. Stuff like “time dependent” aspects of the assembly feature manager. You had to identify feature names, symbols, cursors, menus, etc.

The modeling questions were reasonable. The questions and diagrams were clear and understandable, so that if you got it wrong, it wasn’t a question of interpretation. It made sure you knew how to make sketch relations and use parametrics. To take the test, you had to allot 8 hours, there were 50 multiple choice and a selection of modeling questions. If you could rip through the test at full speed, the whole thing would still take about 2 hours, including the modeling, but many people spent 5+ hours at it. It was not a simple test, but it was for the most part fair. It was an imperfect test, but it tested a wide range of knowledge, and if you passed it, you felt a sense of accomplishment.

This past fall 2007, the test has been redone again. There are several changes. First, they have added a new level for students and beginners called the CSWA (A=associate). For the CSWA, there are only 7 multiple choice questions, and a 3 hour time limit. There are three modeling questions, where you select a mass property in the multiple choice. The difference between a “part modeling” and “advanced part modeling” is simply the number of features, sketches or dimensions. It’s not clear to me why you would bother with “advanced part modeling” on a test for beginners. Also, the certification terminology doesn’t line up with the training terminology. “Advanced part” in training implies something with lofts, sweeps or something complex or “advanced”. On the CSWA, you will not use any advanced features whatsoever.

On the CSWP you also have two COSMOSXpress questions. This seemed kind of odd to me. Why COSMOSXpress? Why not Routing, or ScanTo3D or sheet metal or surfacing? There is even a separate COSMOS specialist test that you can take if you want. I suppose it has to do with the general marketing push that we have all seen on COSMOS stuff. I think the COSMOS questions are out of place on the CSWA and the CSWP, and should be reserved for the COSMOS specialty test, but I get the feeling this is another non-negotiable marketing initiative.

On the CSWP, you have 18 questions, still the 3 hour time limit, half for parts, half for assemblies, and they are all fill in the blank, and all modeling questions (no “word” type questions). If you get the first one wrong, you’re screwed because the questions build on the original part. Half of the questions are based on just making changes to the model dimensions, and reporting the new mass properties.

The CSWP is not significantly more challenging than the CSWA, aside from the number of questions. In fact, on the CSWP you have no questions on drawings!?!? You still have COSMOS questions, but no drawings. You can get through the test using a Rib feature as the most advanced operation.\n\nThe most challenging part of either test is reading the illustrations. They are 3D isometric dimensioned views. I bought the college text book that they took many of the parts for the training classes, certification parts and even Model Mania parts. The text book does things the same way, and many of the parts in the text have dimensions that do not work together, or geometry that is impossible to see, dimensions that are impossible to interpret where they attach, etc. Anyway, the faults of the book spilled over into the test.

In the end, the new test mainly tests your interpretation skills. The modeling that it tests is BASIC at best. Extrude, revolve, cut, fillet. No sheet metal, surfacing, no advanced fillet types, no loft, sweep, deform, indent, multibodies, no nothing that could be considered even remotely advanced or challenging. The knowledge of SolidWorks that it tests is zilch. Someone on the SW Forums pointed out that you could take the CSWP test using Inventor, without even firing up the SolidWorks software.

The new test is a paragon of efficiency. It is completely automated, geared to low quality and a high number of test takers (there’s that “scalability” again!). I’m sure that was part of the goal, along with pushing COSMOS on a captive audience. The CSWA is almost something I could live with, if the COSMOS stuff went away, and you had many more questions covering a wider range of essential concepts. The CSWP is a lame afterthought. I don’t get it. What does it test? The parts are basic, but the illustrations are abombinable.

On the bright side, the cost of the test is coming way down, from $500 to $99. Also, there will be some specialist tests, like sheet metal, COSMOS and surfacing. These tests are only $20. Someone suggested that the tests are really just ways for SW to take more of your money. I don’t think this is true at all. I do think they are a thinly veiled COSMOS marketing ploy.

Rather than just criticizing, here’s what I would do if I were in the thankless position of writing certification exams:

– Make sure the test measures your Skills and Knowledge of SolidWorks software.

– Make sure everybody has to do a mixture of word questions covering concepts and modeling questions covering more than just the basics.

– Test many corners of the software, maybe allowing the user to select questions to leave out – everything – sheet metal, mold tools, surfacing, multibody, shape/deform/flex/indent, loft, sweep, hole wiz, in-context, drawings, inserted parts, patterns, mirror, sketch relations, configurations, etc

– Make sure the questions and illustrations are absolutely clear and unambiguous. The models must be challenging, the illustrations should not be. This must test modeling skill, not interpreting ambiguous illustrations.

– Ask tough relevant questions with clear answers, don’t get into playing word games or trying to be cute (the old test was especially bad in this respect).

– Don’t let the technology used to deliver and grade the test get in the way of writing a challenging, meaningful exam.

Jeremy Luchini is the guy at SW responsible for the test, and I really don’t envy his task. He has the best of intentions. Everybody’s a critic, and I’m sure he’s heard it all.

Anyway, I don’t really expect anything I say here to go anywhere. What’s done is done. You still have to take the test to get on the CSWP list. You need to go to the CSWP event at SWWorld. You need the bragging rights, even if they aren’t worth the 1s and 0s they’re printed on. The next step is to go to the SolidProfessor website and order their new CSWA/CSWP  preparation class. Yes, I wrote it, so this is your cue to jump all over me for advertising. (ttthhhhppppbbbbbbbtt!!)

0 Replies to “SolidWorks Certification Process”

  1. I really hope SW gets it hand burnt in this one..that will force them to focus their resources on 3D MCAD and do more on that they are good at. These marketing folks in SW who hoped to convert some ADSK customers by plugging some letters will learn that it is functionalities that sell their product to somebody using ADSK.

  2. The laywers are smart enough to know Autodesk made some mistakes earlier in its existance. However, there is a question that “dwg” was ever trademarkable since it is derived from pre-existing material (it is the official abbreviation for “drawing” in the standards that Autodesk did goto in determining the file name convention). It’s a bit like McDonalds trying to claim their trademark superseded the family from which they derived their name. These actions by Autodesk needs to be fought against. That said, the “dwg” lawsuit should have never happened. Autodesk’s case against SW does include other factors that I have no opinion about, so I’m not going to critize the overall suit; just the “dwg” matter.

  3. I have a Google blog alert that tells me when my name comes up. Today it pointed me here, as my name was among those listed as having posted on the subject. Apparently, in an earlier revision. Not sure why you took my name out — though I suppose it might be because I’ve actually tried to be quiet on the Autodesk v. SolidWorks suit.

    Let me clarify what this is about: It is just business. Just like Autodesk passing out disparaging handbills at SolidWorks World, this action is about gaining an edge over a competitor. Whether anyone finds it egregious is really immaterial, if, in the end Autodesk is able to show bottom line results, and Carl Bass’s stock options are worth more.

    Is an all out legal war between SW and Autodesk really in the best interest of customers? Of course not.

    But Autodesk is not managed by idiots, and they don’t hire stupid attorneys. They’ve certainly done the math on this one, and have likely concluded that win or lose, they still win.

    If SolidWorks folds up their tent on this, I’m pretty sure it’s not going ultimately be a good thing for you, or any of their other customers.

    ****

    I don’t claim CEOs are idiots. Just that there is a good likelihood that they don’t act with their customer’s best interests in mind or that they are simply misguided. Obviously, as evidenced in what we have all seen in the news over the last couple of weeks, a lot of business people use emotion rather than intellect to make decisions.

    Yes, the first version of this post had your name in it because I remembered your comment on Ralph’s blog, but decided that since it wasn’t on your blog, I wouldn’t saddle you with that burden.

  4. I think that only non-Autodesk users can win anything here. If Autodesk is allowed to tighten it’s control of the dwg file format by establishing trademark rights over it, then non-Autodesk users will suddenly have a harder time interoperating with the de facto 2D CAD file standard, and Autodesk users will have to pay for the artificially inflated value of AutoCAD. Unless the other vendors are able to negotiate a license with Autodesk, which they may have to alllow to avoid being sued for exercising monopoly control over a de facto standard (which they could have if they had their way). So, SW is defending our ability to use the dwg file format. Your work may not require it, but I have to deliver that crap all the time.

    If Autodesk win, Autodesk users lose. By erecting an artificial trademark barrier around “dwg”, Autodesk can continue to charge too much for AutoCAD. Autode$k indeed.

    Obviously, any CAD company is in business to make money, but Autodesk is just nasty about being greedy. They have a history of manipulating the users to extract more cash from them.

    So, SW is using it’s lawyers for the good of keeping .dwg available (thus allowing them to make money on it too), and Autodesk is using their lawyers to enable naked greed. “[Lawyers] are like any other machine: they’re either a benefit or a hazard. If they’re a benefit, it’s not [a] problem.”

    The whole “trademark dressing” thing, I think, is just Autodesk being a bunch of petty jerks. Like they have an exclusive right to orange rectangles. Good grief. Of course, they may not only be petty jerks. This may also be some tactic to stall or otherwise encumber what may be a successful image thing for SW. Who knows, maybe it’s a way of getting cult of Autodesk believers to think that SW just copies Inventor. I had the misfortune of looking at a, IV blog post recently. The blogger was comparing SW09 and IV 09. It was obvious that he knew nothing of SW, but he misunderstood it enough to think that SW has been playing catch-up with IV.

    I am truly amazed that you have no opinion on this law suit. Not that I think you should have seen it the way I do (OK, maybe), I’m just amazed you have no opinion.

    ****

    Dale,

    I don’t have any opinion because I don’t use dwg format much. I’m also rather jaded by the legal process, and think that it rarely has much at all to do with the stated dispute.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.