Model-based continued: Model-Based Definition
After a short celebration, 10 years blogging and 200 posts, now it is time to continue my series related to the future of model-based. So far my introduction and focus on the bigger picture of the term Model-Based has led to various reactions. In particular, related to Model-Based Definition, the topic I am going to discuss in this post. Probably this is the topic where opinions vary the most as it is more close to the classical engineering and manufacturing processes.
What is Model-Based Definition?
There are various definitions of the term Model-Based Definition. Often the term Model-Based Enterprise is used in the same context. Where some people might stop thinking because the terminology is not 100 % aligned, I propose to focus on content. Let’s investigate what it is.
In the classical product lifecycle, a product is first designed for its purpose based on specifications. The product can be simple, purely mechanical or more complex, requiring mechanical design, electronic components, and software to work together. For the first case, I will focus on Model-Based definition, for the second case I recommend to start reading about Model-Based Systems Engineering approaches where the mechanical design is part of a more complex system.
Model-Based Definition for Mechanical Designs – the role of 2D
Historically designs were done on the drawing board in 2D. After the introduction of 2D CAD and later affordable 3D CAD systems at the end of the previous century, companies made a shift from designing in 2D towards 3D. The advantages were clear. A much better understanding of products. Reading a 2D drawing requires special skills and sometimes they were not unambiguous. Therefore, 3D CAD models lead to increased efficiency and quality combined with the potential to reuse and standardize parts or sub-assemblies in a design.
These benefits were not always observed as complementary to the design (the engineering point of view), there was still the need to describe and define how a product needs to be manufactured. The manufacturing definition remained in a set of 2D drawings, and the 2D Drawings were the legal authority describing the product.
An interesting side note observation:
You will still see in industrial machinery companies, a pure EBOM does not exist, as designs were made to target the manufacturing drawings, not the 3D Model, engineering focused, intent. In this type of companies, the discussion EBOM/MBOM is challenging to explain.
Once the 3D Model becomes the authority, the split between design and manufacturing information will create extra work if you keep on creating 2D drawings for manufacturing. It requires non-value added extra work, i.e., reinterpreting 3D data in 2D formats (extra engineering hours) and there is the risk for new errors (interpretations/versioning issues). This non-value added engineering time can add up to over 30 percent of the time spent by engineering. You can find these numbers through the links below this post. I will not be the MBD teacher in this post, I will focus on the business impact.
Model-Based Definition based on 3D
3D PDF Model
The logical step is to use the 3D Model and add manufacturing information attached to the model, through different views. This can be Geometric Dimensioning and Tolerancing information (GF&T), Quality measurement information, Assembly instructions and more, all applied to different views of the model.
Of course here you become dependent on the chosen environments that support the combination of a 3D CAD model combined with annotation views that can be selected in the context of the model. There are existing standards how to annotate a model, find your most practical standard to your industry / Eco-system. Next, most CAD vendors and PLM vendors have their proprietary 3D formats and when you stay within their solution range working with a model-based definition will bring direct benefits, however …..
Model-Based Definition data standards
Every company needs to be able to combine and share information internally with other teams or with partners and suppliers, so a single vendor solution is a utopia. Even if your company has standardized themselves to one system, the next acquisition might be disturbing this dream. Anticipating for openness is crucial and when you start working according to a model-based definition, make sure that at least you have import or export capabilities from within your environment towards model-based definition standards.
The two major standards for model-based definition are 3DPDF and AP242/JT based. Don’t expect these standards to be complete. They will give you a good foundation for your model-based journey and make sure you are part of this journey. (Listen to the CIMdata webinar also listed below)
The Model-Based journey
It took almost 20 years for 3D CAD to become the mainstream for mechanical design. Engineers are now trained in 3D and think in 3D. Now it is time to start the journey to abandon 2D and connect engineering, manufacturing and service more efficient. Similar gains can be expected. Follow the links below this article, here already a quote from an old post by Isha Gupta Ray (Capgemini) related to MBD:
MBE Drivers: The need for consumption of 3D product data by non-engineering departments and the elimination of 2D drawing related rework and costs are driving companies to adopt 3D MBE methods rapidly. DoD predicts that the move away from 2D Drawings and into open and free-to-view 3D MBE documents will reduce the cost of its internal engineering activities by up to 30%, reduce the scrap and rework it currently deals with from its supply channel by nearly 20% and improves supplier response times by up to 50%.
Conclusion
Model-Based Definition is not as challenging as becoming a Model-Driven enterprise, that I described in my introduction post to this theme. It is a first step to challenge or energize your company to become a digital enterprise, as sharing between engineering and manufacturing needs to be orchestrated, even with your external parties. It is easy to do nothing and to wait till your company is pushed or pushed out, which would cause extra stress (or relieve forever). For me Model-Based Definition is a first (baby) step towards a digital enterprise, warming-up your company to change a look at your data in a different way. Next when you combine parameters and simulation to your models, you will make the next step towards a model-driven digital enterprise.
Below a selection of links related to the theme of Model-Based Definition. If you feel I missed some crucial links, please provide them through the comments section of this post, and I will add them to the post if relevant.
Tech-Clarity: The How-to Guide for Adopting Model Based Definition (MBD)
Action Engineering: Articles, Blog plus training
Engineering.com: How Model-Based Definition Can Fix Your CAD Models
Lifecycle Insights: Quantifying the value of Model-Based definitions
CIMdata: Webinar on Model-Based Definition and Standards
Capgemini: Model-Based Enterprise with 3D PDF
if you want to learn more in-depth the advanced usage and potential of MBD, try to understand:
Jos Voskuil's Weblog 
1 comment
Comments feed for this article
June 7, 2018 at 10:55 am
paulvdree
Jos,
Here is a fellow Dutchman and avid reader of your blog as well as Oleg’s. Thanks for taking the time to share your insights with the world.
I have been following the MBD debate. Especially now that ASML is going to push PMI from now on it is going to be very real for suppliers in all tiers of the ASML supply chain.
The conclusion that I have made so far about MBD/ PMI is that is a case of “solving the wrong problem”.
I could write this all up in my own words, but will just point you to the source of my insights:
http://www.tecnetinc.com/newsletter%20topics%20original.html
This is a link to all the topics that Mr. Joe Brouwer has covered.
More specifically on MBD/ PMI:
http://www.tecnetinc.com/PMI%20vs%20AID.html
http://tecnetinc.com/The%20Death%20of%20the%20Drawing.html
The takeaways (combination of Mr. Brouwer’s and my own insights):
– A drawing from a 3D CAD system is not a drawing, but an AID (Associated Information Document);
– The 3D model has always been the master, the AID is derived from the 3D source (isometric and additional projections). Ever since the arrival of 3D CAD, no drafting is done anymore;
– The AID is and must be clear, concise and unambigious. “Drafting” standards insure this and they are understood worldwide;
– The AID must not be cluttered with dimensions that are already captured in a general tolerance (i.e. +/- 0,2 mm for general machining). This does not add value (the 3D model is the master/ template). ASML got this very right with their General Standard for MDD (Minimal Dimensioned Drawing). The TPD (Technical Product Documentation) always contains the STEP file and the AID (drawing). For capturing the full contour tolerance, ASML uses the “profile of a surface” tolerance, which works very well.
– In the product development/ engineering environment, working with only the 3D model with all GD&T annotations in it, is terrible (personal experience). It is very difficult to see on which geometry datums and tolerances are attached. Also for example for reviewing, markup etc. (lead engineer, purchase, suppliers etc.) it is creating more much problems than it solves.
– The way people communicate and consume information and which format is best suited for that should always be the main driver for innovations. It seems with MBD/ PMI the human interface is overlooked. I have worked with Catia PMI on the F-35 project at Fokker and this proved to be an absolute disaster – people are snipping images from the 3D Catia model and printing them on paper! It’s even worse when your parts are physically large (as in aerospace).
– The “efficiency” focus on saving time “drawing” is overrated. As Mr. Brouwer points out somewhere, it is much more important to make sure that properly designed and reviewed parts make it to the outside world. Senior reviewing is critically important – peer reviewing (if even that) is common practice in many companies nowadays. The efficiency factor may come into play where products are CTO (Configured To Order) – especially in products with a large number of relatively low cost mono parts (e.g. sheet metal constructions). In this area there are possibilities to fully automate and integrate the process, as everything can be parameterized, although this carries large upfront and may create a new level of vendor lock in. In my experience anyway, in the High Tech/ Aerospace/ Medical etc. the focus should always be on quality, not ultimate efficiency.
In conclusion, I think MBD/ PMI fits into the larger MBSE paradigm and “digital PLM”. Practically, I can only see this working when the human interfaces are all eliminated and everything is end-to end data driven. For now, I am not putting my money on eliminating documentation in any form.
I am involved in a PLM initiative in a mid-sized company (~350 FTE). We have Aras Innovator running in a test environment. The implementation (if we get that far) will be to first get our document and change management in place, filling the PLM with released documents in an immutable data format (e.g. STEP, PDF etc.). This will likely eliminate the bulk of our troubles. After that, we’ll be looking to integrate the authoring tools (MCAD/ ECAD/ Office etc.). Only after that MBSE and further digitalization appear on the horizon.
(note: re-post – 1st didn’t appear)
Paul thanks for your extensive response. Very well appreciated as you are echoing an opinion, but using arguments what are to your experience misconceptions and challenges. I know Mr. Brouwers opinion very well as he responds mostly by hyperlinks to posts in the context of PLM or this topic. I prefer more a discussion and as you are a fellow Dutchman, i am thinking of inviting you and another fellow Dutchman who is a big supporter for MBD/MDE to have a joint debate, potentially somewhere in the summer months, when at least my traveling is less intensive. Then, let’s publish or share this discussion to a wider audience, either in a blog/vlog or as part of an upcoming PLM conference.
Related to your second point, the PLM initiative you are mentioning, we can have a separate debate. As you describe it currently in a very document-centric approach I would call it more a previous century PDM implementation with some PLM aspects. Current best practices are already based on an item-centric backbone and for sure in the next 10 years we will learn (and develop) best practices for a model-driven enterprise, where digital continuity is the only way to remain efficient and manage the complexity of products.
I will invite you to elaborate on the points above through my email account.
Thanks again for your contribution, looking forward to the follow-up.
Best regards, Jos
LikeLiked by 1 person