engineering-matters - Latest Comments

Re: Digital Sketching and the Engineering Notebook

James Self — Mon, 14 May 2012 11:16:12 -0000

Many thanks for your words on my post on Core77. I'm becoming more interested in the ways in which designers from different backgrounds, including more, how to put it, 'engineering' orientated designers, approach and employ designerly tools.
James.

Re: Can SDM Be Effective without CAD PDM?

Keith Meintjes — Tue, 08 May 2012 16:09:42 -0000

Chad,

What you say is sort of true for a single component, but for most cases you have oversimplified (or understated) the problem.

In a PDM system, parts are organized into assemblies, and then these assemblies exist in a tree-like organization of systems and subsystems, the product structure. There is usually not a single item in the CAD PDM product structure that maps precisely to what is needed for a CAE model.

The first task is, often, to build a CAE configuration, or BOM, that specifies the required CAD information. As a further complication, this BOM depends on the CAE discipline: Structures, thermal, aero, etc.

Managing the relation of these CAE BOMS to the PDM product structure is a complex task, and is one of the reasons a generic PDM system cannot easily be applied to manage CAE data. This all occurs before your first step, the abstraction of CAD into a CAE representation.

Re: Brave New World: PLM without PDM

Mark Mitchell — Fri, 04 May 2012 22:40:52 -0000

Chad, this debate has been raging for years. As one of those people who have been around since the terms PDM and PLM were first coined I have been bemused the debates' ebb and flow. Like many other terms relating computing software / solutions they are entwined with the evolution of computer hardware and consequently software companies trying to differentiate themselves from their competition. 25-30 years ago the nature of computing and computing tools were very different. Some terms and acronyms come and go, others like PDM and PLM persist because they adequately describe a universally accepted business function/process - God knows we don't need more - there are now 33 definitions of PDM and 23 definitions of PLM!! I digress. PDM came before PLM - the understanding is in the history - not recent but "ancient" history. Now moving forward - while PDM came out of activities involving large amounts of data e.g. engineering. The word Product has evolved to encompass the "electronic file" and therefore the use of the "PDM" term has moved into a wide variety of industries. Likewise PLM simply allowed companies to consider trying to control more than just the product data - SR/CAD/CAM/CAE/DM/Workflow etc.etc. and it to has evolved and is now relevant to more industries.
In the end - "horses for courses".

Re: Is CAD Still Disruption Ready?

Jeff Waters — Fri, 04 May 2012 09:00:05 -0000

I'm inclined to think that every company should give serious thought to looking at their offerings as a platform rather than pure product.

Re: Brave New World: PLM without PDM

Hakan Karden — Wed, 02 May 2012 01:52:31 -0000

Hi George,
this is good. PLM is about federation and then there has to be something in common. We propose standards like STEP and PLCS. For external collaboration we need to add federation of identities as well.
Wrt CAD - the CAE domains has some pretty large files too.

Re: Brave New World: PLM without PDM

George Hielscher — Tue, 01 May 2012 15:20:09 -0000

PLM is about the data. The "PLM database" is the virtual mega database that synergistically combines the product, process, resource and purchasing (supply) data. The "PLM database" is probably a combination of multiple databases with links between them. To me, without the links between the product, process and resource data, it's not really PLM.

On top of the PLM data, we have PDM (including CAD), process, ERP and SCM applications (among others) - probably several applications in each of the traditional categories. But I don't know what a PLM application would be.

A key differentiator of PLM over the previous generation of systems is the synergistic use of product, process and requirements data together. Applications in the PLM super system can take advantage of data from other domains to establish the context of the user's inputs/commands/operations, and can apply more "design intent" intelligence to the operations. For example, when building the product data, the PDM or CAD system can use the current or related process and requirements data to intelligently automate the operations and fill in some of the product data. (Note: I called this the PLM super system because it is made up of applications from multiple software domains, and possibly multiple vendors.)

I think the relation of the PDM (product) data to the CAD system is not inherent, but is probably necessary due to the database performance issues that CAD systems encounter for large designs. Until the database organization of CAD data becomes more standard (between different CAD systems), the product data will need to be designed (optimized) around the CAD data.

To me, the goal of PLM is to bring together the data previously in separate systems to have a more intelligent database that can optimize and automate the product design, manufacturing and operation work. For PLM to be something new, it needs to be a step beyond PDM + Process + ERP + SCM.

Re: Brave New World: PLM without PDM

Rory Granros — Tue, 01 May 2012 14:36:17 -0000

Chad,

Lately, I have been working with supply chain organizations and my view of PLM is a less rigid. Based on the product complexity and your companies role in the process (ATO of a top level BOM), the need for a complex PDM solution will vary. If you have 1000s of parts and multi-level BOMs, you will benefit from PDM.

I believe that a hybrid model will evolve for PLM. This maybe a process industry variant of PLM. As more companies expand globally, you still need to ensure quality and safety of products but the products can quite different and local business vary different.

- home country with most of R&D will create concepts and probably mfg
- as you move into 2nd markets, the products are tailored to the market and reqional mfg is done
- as you move into lower tier markets, you may outsource R&D and some or all the mfg. Regional or corporate quality and compliance and brand managers will still own quality, compliance and branding.

You will have a global process to move from the home market to the lower market. With each region and local market, the processes need to mirror the local processes. If you outsource R&D, the level of detail is much less than when you own R&D.

In these environments, a hybrid approch with global quality, compliance and brand management with local exection will be required.

rory

Re: Brave New World: PLM without PDM

Hakan Karden — Tue, 01 May 2012 13:49:07 -0000

Hi Chad,

Eurostep's Share-A-space product is PLM. It is even PLM in the extended/virtual enterprise. PDM is typically implemented as a Team Data Management application or document management solution. A PLM application should be fine granular, standards based and made independent of IT and Processes. Then it stand the best chance to live as along as the product itself, from concept to disposal. And data can be moved around, exchanged and shared.

STEP, PLCS and AP233 are great candidates as the base for this canonical PLM model. Collaboration and data sharing should be part of a PLM solution. Looking at any product today there is much collaboration before and after it enters the market. This is another reason to be agnostic to technology and processes when it comes describing the data in PLM since a supply chain should not be forced to use the same SW tools.

To figure out what is needed. start looking at extended enterprise and the complete product life cycle including archiving. Then use this requirement and look inwards to drive the sub parts and how they need to interface to be part of real PLM.

Re: Brave New World: PLM without PDM

David_Chisholm — Tue, 01 May 2012 11:42:22 -0000

Hi Chad,

There has been quite a change and business success since you last did your review.

You may wish to get an update?

David

Re: Brave New World: PLM without PDM

Chad Jackson — Tue, 01 May 2012 10:40:07 -0000

Hey Rory. Thanks for commenting. Hadn't heard from you in a while.

I think you have a valid point regarding 'flexibility' and 'light integrations.' You certainly could integrate more readily if there were light touches between systems.

Regarding teams of less than 5 engineers not needing PDM, I actually disagree. I think it has far more to do with the complexity of the models that are being developed than the number of engineers working together. Trying to manually manage the 'file' configuration of a CAD assembly with over 1,000 parts is practically impossible.

I also disagree with you that a top level PLM process controls the global rollout. In my experience, you have different divisions if not even difference offices deploying PLM for their own initiatives. It comes down to the classic 'centralized control' vs. 'local empowerment' argument. I don't think all manufacturers follow that centralized control model from an IT perspective.

Re: Brave New World: PLM without PDM

Chad Jackson — Tue, 01 May 2012 10:34:40 -0000

Hey Dave. Good to hear from you. Thanks for commenting.

I have reviewed Actify's Centro product. You can find the post here: http://www.engineering-matters.... I think it's a pretty fascinating product and could add a lot of value for engineers. I think you have a point here. In the post above, I mentioned that something like Inforbix or alcove9 could be tech that would assist in the 'light integrations' in such an ecosystem. But something like Actify's Centro could fill that same role. My question would be if Centro could find data to the same extent that Inforbix or alcove9 could.

Re: Brave New World: PLM without PDM

Chad Jackson — Tue, 01 May 2012 10:31:18 -0000

Hey Jay. Thanks for commenting. I have looked at Synergis, but haven't had an opportunity to write about them yet. It's in my queue.

I think their capabilities do address some of the multi-PDM concerns. As a result, it would probably be easier to do 'light integrations' across such an IT ecosystem.

Re: Brave New World: PLM without PDM

Jay Lovelady — Tue, 01 May 2012 09:57:23 -0000

I also wonder if you have looked at PDM systems that are not CAD specific such as Synergis's Adept? and how it would fit into your scheme.

Re: Brave New World: PLM without PDM

David_Chisholm — Tue, 01 May 2012 04:40:58 -0000

Hi Chad, Good article. Have you had a look at Actify Inc's Centro? and where do you see it fitting into the mix?

Re: Brave New World: PLM without PDM

rory_granros — Mon, 30 Apr 2012 21:02:57 -0000

For different types, level of PLM/PDM maturity and sizes of
companies, tightly integrated PDM/PLM, loosely bound PDM/PLM, document based
PLM and a federated PLM/PLM/PLM all have roles to play.

·
The value of a tightly integrated PLM/PDM is
well documented.

·
Loosely bound PLM and PDM with well
documented/controlled “critical control points” can all the PDM and PLM’s
configuration and business process to have some level of independence. This can
allow additional business flexibility without impacting the other system and
reduce the time and cost of upgrades.

·
For companies without a significant R&D or
engineering, like banks or insurance companies, document based PLM makes more
sense. Even or many small mfg companies with less than 5 engineers or
formulators, and one location, PDM are often overkill.

·
As I talk with more companies, I wonder if a
federated approach to PLM and PDM will evolve.
In reality, a top level PLM process controls the global rollout. For
each region and maybe department with the region, they could benefit from local
PLM processes. Supporting each of level of PLM, a federated PDM strategy could
more closely mirror the real world. As you move down the “as” versions of the
BOMs, allowing each “as” to have more autonomy while provide a global version
that can be federated would be beneficial.

As PLM processes mature but need to evolve at a faster pace,
a federated model could be right balance between consistency, compliance and
speed.
The view of process PLM guy.
rory

Re: Oracle’s Fourth Evolution of PLM?

Chad Jackson — Thu, 26 Apr 2012 11:40:45 -0000

And the circle is closed. C'mon, it's only 3600 words!

In all seriousness, I think you have a valid point. If the story here is mostly about access, then yes. There are other solutions that are more granular and easier to deploy. Agreed.

If part of the story is about duplicated data (here I mean objects and their metadata, not duplicated files based data), then I see some value in an MDM hub that acts as the source of truth and pushes it back out. But I think those cases tend to be limited to much larger companies.

I do see some value in executing those processes that span multiple enterprise systems. However, like I said, I'm not sure building integrations one at a time is the most efficient and flexible manner to go after it. Maybe someone from Oracle's Agile team can comment on that aspect.

Great points Chris. Thanks for contributing.

Re: Is PDM Disruption Ready?

Dave — Thu, 26 Apr 2012 11:34:41 -0000

Hi Chad,

I think there are couple of vectors from which providers come. On the one hand, there are cross-organizational product suites from Oracle, IBM, SAP, and Microsoft, that have substantial workflow engines in them. Implementing (and extending) those, rather than creating an island, seems to be one potential strategy. I'd also throw in the workflow products attached to the ECM solutions such as EMC/Documentum, OpenText, FileNet, etc.

The other is independent workflow providers such as Nintex, K2, etc. that either augment the standard capability in one of the stacks (i.e. Nintex to SharePoint), or stand entirely on their own regardless of what the underlying architecture is (i.e. K2). To me, extending workflow from within a PLM system to cross organizations doesn't make sense when the enterprise as a whole is going a different direction from a workflow implementation strategy perspective.

Dave

Re: Oracle’s Fourth Evolution of PLM?

Chris Kelley — Thu, 26 Apr 2012 11:23:35 -0000

I'm not sure how I ever made it through such a long post since Google has made me so dumb and inattentive ;-).

Two quick reactions:

1 - Not sure this is really a 4th revolution. PLM Vendors have had the concept of PLM as extended enterprise application. Most of the "collaboration" examples you see always include an OEM and a supplier working on some widget.

2 - They seem to be coming at the same problem (access to information / duplicate information) as companies like Inforbix (free plug for Oleg ;-), but coming from a diametrically opposed starting point. Oracle wants to add more structure to solve the problem of too much structure (like certain governments deciding the solution to their debt problem is more debt)....maybe we need a something less complex and structured to really get somewhere?

Re: Is CAD Still Disruption Ready?

Joe Brouwer — Wed, 25 Apr 2012 07:10:31 -0000

Hi Sri,

Your request would only work if we all used the same program. The problem in the industry is that we have no standard.

In the past I could walk into a job with a mechanical pencil, triangles, compass and scale and start working. My only requirement was my experience. I could instantly work on someone else's design totally understanding the design intent.

Today the prerequisite is the CAD system. You can be the most experience Flight Deck controls designer and if you are not an experienced Catia user you do not get the job. As I was out of the working engineering world (CAD reseller provide engineering services). I found that the designed drafter has been replaced with the engineer. I feel engineering quality has dropped for that reason.

I will be writing an article on how the high end systems have corrupted engineering quality. You have to be on the outside of the industry to see this. We truly need a standard to get the engineering cost down.

Of course I could go on and on.

Please visit my website and peruse some of my newsletters about the state of CAD and CAM.

http://tecnetinc.com/newslette...

Re: Is CAD Still Disruption Ready?

Joe Brouwer — Wed, 25 Apr 2012 06:57:30 -0000

Sorry for not getting back to you.

Designing on CAD is more than manipulating graphics. Most of my time is checking fits and selecting hardware. Every entity you use has to be scrutinized. You can only work at a certain speed. You may ponder on a problem for days doing different iterations of your design. You may work for hours on how to make two surfaces tangent only to find that it is beyond your surfacing capabilities.

Years ago voice command was implemented in CADKEY. It was a novelty.

So with 30 years in 3D CAD the mouse and keyboard offer enough flexibility to get the job done. The Spaceball provided graphics control for those programs that did not offer the standard manipulation of the screen. I have used it with one program.

Most of the programs allow you to customize your settings. The middle mouse button rotates the screen. I use that setting if the program allows it. But I took a job as sales manager and AE at a company that was going to represent Autodesk manufacturing products. I found that I could not change the middle button to rotate, it was fixed to pan. I laughed and said "what do you expect it's Autocad". That job didn't last and I decided to learn SW. Of course, I set the middle button to rotate. With SW as well as Inventor they use the Pro/E paradigm where everything is a sketch. All of a sudden I found myself setting the middle button to Pan. When you have a program that utilized 2D sketching you want to pan not rotate. I suppose the spaceball would be a good option there. I have never worked with that paradigm, much more than building a few parts and was a bit surprised with all the hoops I had to jump through.

I suppose I could go on. But we really don't need another interface beside the Mouse and Keyboard at least on our Laptops and Desktops.

Now lets go to tablets.

I have an Android tablet and used "Go to my PC" to access my CAD program, it came in and was very clunky utilizing the touch controls. I even plugged in a mouse and keyboard and it was basically useless. Luckily, my business is not conducive to tablets at least at this time.

I am a bit excited about Windows 8.

Re: Is CAD Still Disruption Ready?

George Hielscher — Tue, 24 Apr 2012 17:29:59 -0000

CAD is getting old and boring, and more complicated at the same time. Many of the old school CAD vendors are putting emphasis on pushing/pulling points on complex surfaces, without much respect for designing with precision. It seems that some of the most simple things are hard these days - like positioning one object precisely relative to another object (but it's easy to drag the object to an approximate position).

What's the solution to old and boring? Let the system do the boring stuff for you. For real world CAD models, we often position "features" (objects, components, parts, or whatever we end up calling them). With better capture of design intent (what the feature is used for or represents), the system can make better decisions about how to place the feature. And not coincidentally, that is the same information about the part that we need for PLM.

So with semi-automatic placement and design of each feature or part, the imprecise drag-drop of objects can be correct more often. And since the system knows what the feature/part does, it also knows how it needs to be connected (a combination of mechanical, electrical, and possibly fluidic/flow/process connections). Initial connections can be proposed (maybe assuming the designer is thinking in the order of connections), and the designer can remove or adjust connections. The discipline of the designer, as well as specific characteristics of the parts and features, tell the CAD system what the designer is (most likely) trying to do.

Side note: I think 99% of real world design involves the placement and assembly of existing parts or already known part features (hole, planar face, mating surface). Only 1% of design involves design of new surfaces or shapes. CAD systems should be providing a lot more design intelligence when placing parts and features, and organizing assemblies.

I know there are some CAD systems that are industry specific that apply a little intelligence in placing and assembling parts and features, but they are not the prominent CAD systems, and they are not that smart at placement. Why? Because they lack an understanding of the part or feature - the function and connections. And this greater understanding of the function and connections is what we get with PLM.

So I think the next, biggest disruption in CAD will come from the increased design automation that can be applied from the PLM knowledge of the current design, parts and features in this design, and the designer (and his organization).

And why has this not happened yet? CAD providers are too busy just trying to fit into a PLM environment. And this kind of intelligent CAD takes a big PLM database and some heavy-duty processing power - too much for the workstation used by most designers (often a laptop or remote PC). So in comes "CAD on the Cloud" to the rescue.

"CAD on the Cloud" can provide the PLM database access and occasional heavy compute power to allow intelligent design automation. (Note: this implies PLM on the cloud too.) Then the CAD "workstation" can be less powerful, connected by WiFi, and it manages primarily the user interface and graphics. Here is where the powerful touchscreen tablet can fit in. (Generally, a tablet does little without the internet or the cloud.) A powerful tablet could manage the graphics (assuming the right amount and complexity of the geometric model is downloaded into the tablet), and the cloud application server can do the heavier portion of the intelligent part/feature placement. And although the touchscreen is not so precise, intelligent positioning (snapping) of parts according to the connections can provide the precision in many cases.

So tablets and touch screens can easily be in the future of CAD, and voice commands can work better with the added context from the PLM information. Mobile devices can work if "CAD on the Cloud" is combined well with PLM on the cloud. I know that these ideas are "way out there", but mobile and tablet technologies are moving fast, and we may see technological advances quickly if PLM ever gets straightened out.

[Note: This comment is phase 1 of the CAD disruption that I foresee. I hope to comment on Phase 2 CAD disruption later if I get time. It is even more of a change, and even more "out there".]

Re: Calculations and the Engineering Notebook

Chad Jackson — Tue, 24 Apr 2012 15:08:40 -0000

Pretty amazing how policies have changed. I think you've got something there with regards to Maple or MathCAD types of tools. Although they seem to be missing other capabilities like sketching, notetakeing and the like. I need to take a hard look at something like that next. Thanks for weighing in.

Re: Is PDM Disruption Ready?

Chad Jackson — Tue, 24 Apr 2012 15:07:17 -0000

Thanks for commenting Dave. You make some very valid points.

Specifically, your comment about workflows sitting over the top of several organizations and their cross-organizational processes is what caught my eye. I think you're correct. What sort of software providers do you see filling that gap? Are you just talking about the huge ones or some that offer process specific solutions? And are they closely or loosely connecting to product data and information?

Thanks again for contributing. Good points.

Re: Is PDM Disruption Ready?

Dave — Tue, 24 Apr 2012 11:35:39 -0000

Chad,

Very interesting topic. It reasonates more than a little bit with me. Its always been my view that much of what makes up PLM solutions isn't the application logic that composes most "PLM apps", but rather that the whole idea of PLM as a solution "stack".

That idea IMHO dates back to when there really wasn't standard infrastructure, so vendors bundled in databases, workflow, messaging, web servers, etc. because that was seen as the only way to be able to deliver the apps. The world has changed, however. For instance, companies now have workflow engines that can span the enterprise; why would they want to create an "exception" sitting over in engineering? Then there is the whole issue of who owns the data - ERP, or PLM? We won't go there, but I think we all know the answer.

You can trace this line of thinking through the entire PLM stack, execpt for two things - the application logic itself, and....PDM.

To me, it makes a huge amount of sense that companies have gotten "smart" about their IT architecture decisions, and so the vendor exec's experience seems to me to be a reflection of that reality.

Great post.

Dave

Re: Tetra4D: Leveraging the Adobe Ecosystem

Mark Guthrie — Tue, 24 Apr 2012 07:34:45 -0000

Excellent article. I appreciate the depth you went into.