Alibre Reacts to Synchonous Technology
Ever since Siemens released their teaser video on the "breakthrough" Synchronous Technology, I have been talking to a few CAD vendors, asking them what they think about the growing trend toward direct modeling.
Greg Milliken, CEO of Alibre, was kind enough to share his views with me. I asked him if Alibre was pursuing development of Direct Modeling techniques as well. He replied that they were and pointed out that version 10 of Alibre Design already had some direct editing capabilities. Indeed, it has. But I believe it needs to be improved to be considered comparable to the advanced technologies of SpaceClaim and Siemens. For example, if you push a fillet to into a solid, the adjacent faces remain as they are and you end up with a jagged fillet. However, I am pretty sure Alibre Design 11 will have far more improved and sophisticated direct editing capabilities.
According to Greg, "We consider it [Direct Editing] a “feature” though rather than a product and believe the future is in hybrid systems with both parametric and direct editing capabilities. The folks at SpaceClaim, Kubotek and CoCreate/PTC staked too much in one approach and I feel over time they will basically lose any differentiation as all parametric, history-based products adopt direct editing toolsets."
Now thats an interesting way to look at it. Actually, I am of the opinion that the opposite may happen. As feature recognition technologies improve to a point that they can capture design intent, Parametrics may lose out to Direct Editing. For example, currently in SpaceClaim, if open a dumb solid model of a flange having 8 holes of equal radius, SpaceClaim does not recognize that those holes are actually a polar array. If you pick one of the holes and increase their radius, the rest remain as they are. If will be only a matter of time when SpaceClaim and other Direct Editing software will be able to recognize the holes as a polar array and treat them as such. I am not sure whether the Synchronous Technology from Siemens is already close to doing this.
Greg had something quite interesting to say regarding a post on this blog, "Regarding a comment in your blog titled 'SpaceClaim Reacts to Synchronous Technology,' in my opinion, contrary to this adding a zero to SpaceClaim’s value, I think it decreases their value significantly since Siemens announcement, our addition of direct editing capabilities, and essentially all other vendors doing the same, or imminently planning to, means they have little to differentiate themselves technologically that can’t be replicated by most vendors in a relatively short period of time."
Fascinating!! So if Greg is right (and he very well may be) then all the noise about SpaceClaim being eventually bought out by Autodesk or SolidWorks may very well come to an end. If the venture capitalists funding SpaceClaim didn't sell earlier, there is probably no reason why they would sell now.
We are living in interesting times.
Greg Milliken, CEO of Alibre, was kind enough to share his views with me. I asked him if Alibre was pursuing development of Direct Modeling techniques as well. He replied that they were and pointed out that version 10 of Alibre Design already had some direct editing capabilities. Indeed, it has. But I believe it needs to be improved to be considered comparable to the advanced technologies of SpaceClaim and Siemens. For example, if you push a fillet to into a solid, the adjacent faces remain as they are and you end up with a jagged fillet. However, I am pretty sure Alibre Design 11 will have far more improved and sophisticated direct editing capabilities.
According to Greg, "We consider it [Direct Editing] a “feature” though rather than a product and believe the future is in hybrid systems with both parametric and direct editing capabilities. The folks at SpaceClaim, Kubotek and CoCreate/PTC staked too much in one approach and I feel over time they will basically lose any differentiation as all parametric, history-based products adopt direct editing toolsets."
Now thats an interesting way to look at it. Actually, I am of the opinion that the opposite may happen. As feature recognition technologies improve to a point that they can capture design intent, Parametrics may lose out to Direct Editing. For example, currently in SpaceClaim, if open a dumb solid model of a flange having 8 holes of equal radius, SpaceClaim does not recognize that those holes are actually a polar array. If you pick one of the holes and increase their radius, the rest remain as they are. If will be only a matter of time when SpaceClaim and other Direct Editing software will be able to recognize the holes as a polar array and treat them as such. I am not sure whether the Synchronous Technology from Siemens is already close to doing this.
Greg had something quite interesting to say regarding a post on this blog, "Regarding a comment in your blog titled 'SpaceClaim Reacts to Synchronous Technology,' in my opinion, contrary to this adding a zero to SpaceClaim’s value, I think it decreases their value significantly since Siemens announcement, our addition of direct editing capabilities, and essentially all other vendors doing the same, or imminently planning to, means they have little to differentiate themselves technologically that can’t be replicated by most vendors in a relatively short period of time."
Fascinating!! So if Greg is right (and he very well may be) then all the noise about SpaceClaim being eventually bought out by Autodesk or SolidWorks may very well come to an end. If the venture capitalists funding SpaceClaim didn't sell earlier, there is probably no reason why they would sell now.
We are living in interesting times.
posted by Deelip Menezes at
8:12 PM
8 Comments:
I'm not sure I want your direct editing future - it relies on the software "automagically" discerning design intent, while in parametric CAD, the design intent is explicit.
Think of how well MS's autocorrect features work - I always turn them off, because, for example, MS Word always thinks "HSA" (head stack assembly) is "has". So what do you do when your CAD software mis-interprets your design intent?
By
Anonymous, At
10:52 PM, May 01, 2008
Anonymous: "So what do you do when your CAD software mis-interprets your design intent?
"
It may very well happen that a face may be "recognized" as part of more than one feature. The user may be given the freedom to "choose" the design intent in that case. I know it sounds weird, but so did solid modeling when people were drawing with pencil and paper. Untimately, it's upto each CAD vendor to make their software intuitive enough, and at the same time flexible enough, for the end user to not end up tearing out his hair.
For example, take the ribbon implementation by Microsoft and Autodesk. Microsoft literally stuffed it down our throats, whereas Autodesk gave us an option to use it or not.
By
Deelip Menezes, At
11:12 PM, May 01, 2008
For simple cases, like your polar array example, design intent might be easily identified by the non-parametric CAD system, but there are just too many examples where desing intent simply cannot be inferred. One that immediately springs to mind is when features are related to each through equations - how in the world can a CAD system infer all the possible driving equations? For that matter, how can CAD system know what the designer intended in all cases? Impossible. No matter how many options you can present to the user, it is guesswork without actually knowing what the designer was thinking.
I believe tightly coupled hybrid systems are the future of mechanical design.
By
Anonymous, At
12:09 AM, May 02, 2008
Anonymous: "I believe tightly coupled hybrid systems are the future of mechanical design."
You are probably right. However, I am just a bit curious here. In your opinion, is it absolutely necessary to tangle a model with equations, thereby making it difficult for someone else to modify it later on. We have that exact problem in programming. Unless we heavily comment our source code, any complex algorithm is going to be very difficult for another programmer to debug or modify.
By
Deelip Menezes, At
12:21 AM, May 02, 2008
Anonymous, I think you are incorrectly lumping Synchronous Technology with various history-less CAD systems that have been around for a while. It is certainly not that. In Synchronous Technology you can create equational dimensions which relate two features very easily. You can make it as "parametric" and "dimension driven" as you like. The primary thing to remember is that you don't have to suffer a linear-ordered regeneration. But there is no loss in your ability to build intelligence into the model. I'd suggest you read the whitepaper by Ken Versprille on this subject. I believe you can find it at www.solidedge.com.
By
Anonymous, At
1:19 AM, May 02, 2008
In the case you mention, it is easy to simply remove the fillet with our remove tool and then reapply it as a fully-editable parametric feature. There are many ways to “skin the cat” so to speak. Our philosophy has always been to utilize development resources as efficiently as possible, so as we enhance direct editing or other aspects of our product, rather than spending lots of time trying to perfect a particular, and potentially uncommon case, such as “pushing” a fillet until it disappears or cuts into the model in a manner probably not intended by the designer, we’d rather use the resources on other more commonly used features and enhancements. Is having to remove the fillet and simply reapply it, versus using a fancy “Gizmo” to do something you probably won’t do very often, worth thousands of extra dollars? We say no, but of course, if you spent the dough creating the Gizmo you probably think differently. Either way, we’re always focused on improving things, and listening to customers to help us set the priorities. Maybe we'll decide a Gizmo is the way to go and put a junior developer on it. ;-)
Anyway, I like your healthy skepticism.
And one last note, while we’re all about making the product better and believe strong technology is critical to success, we think success is more closely tied to the long-term viability of the business model. Essentially all successful companies sustain themselves based on the model versus the technology alone. Consider Autodesk: AutoCAD has a lot of nice features, but rather than elegant technology, it was the channel and third-party following that really cemented their position and on which they still primarily rely to drive their revenue. And PTC, while they did advance a new modeling paradigm, their aggressive direct sales force was fundamental to driving their growth. Had they fumbled their sales model I expect all the technology in the world wouldn’t have resulted in the success they achieved. Those Halcyon days are long past, unfortunately for them, because they missed the Autodesk-like VAR model of SolidWorks. Recall all the negative things PTC tried to say about SolidWorks 1995 that were likely true back then: it couldn’t model large assemblies, it couldn’t do surfaces, etc., etc. – sounds like what some of the SolidWorks diehards say about Alibre today. But no one really listened because what the market really wanted was an affordable, easier to use solution that ran on their PC. They were also advantaged by discontent in Autodesk’s channel with Autodesk’s lip service to provide a real 3D solution, so a number of Autodesk's better VARs focused on the mechanical manufacturing market jumped ship. The examples go on and on: Dell, Microsoft, Google…
So I ask, who is doing something different with respect to the model? If you can’t find anything different there, then it’s just a “me too” effort that is somewhat late to the party, albeit one that deserves technical respect. Who really thinks “Synchronous Technology” is suddenly going to propel Solid Edge in any significant way relative to SolidWorks?
By
Anonymous, At
2:09 AM, May 02, 2008
I get a kick out of the user comments. They're mostly from people who are expert CAD users.
This is kind of like a bunch of race car drivers saying they don't like a new car because it has an automatic transmission.
I'm in awe of how good some expert CAD users are. But not all CAD users are experts.
Let's put this in perspective. Aberdeen Group reports that the number one challenge to design reuse is this: Model modification requires expert CAD knowledge.
By
Anonymous, At
5:56 PM, May 04, 2008
It requires more than an expert user, Evan. It requires time... lots of time! That's time that machining job shops simply don't have and that their customers don't want to pay for.
The fight is about expert users who don't want to lose the "control" they think they have a right to have. The fact is that they are going to lose this "control" and it’s only a matter of time before it happens.
UG started as a CAM system and I think that to this day this still helps give Siemens/UGS a serious edge over their competition.
Jon Banquer
San Diego, CA
By
Jon Banquer, At
8:32 PM, May 04, 2008
Post a Comment
Subscribe to Post Comments [Atom]
<< Home