3D & PLM: A Foundation for World-Class Innovation
David Prawel, Longview Advisors Inc.

If a picture is worth a thousand words, than a 3D picture is worth a million.

World-class innovation demands excellence in many things – product design, change management, elimination of waste and integration of design and analysis to mention a few – all of which are critical to global competitiveness. From nuclear medicine to architecture, manufacturing to computer games, petrochemicals to weather prediction, 3D software technology is changing the way we communicate and innovate. 3D software, supported by PLM processes and technologies, provides a foundation upon which manufacturing businesses can be transformed and innovation dramatically accelerated.

3D provides a lot of added value. The amount and type of information that can be shared in a 3D model is so much richer than in 2D. For design, engineering, analysis, viewing, simulation and countless other applications, 3D models offer unprecedented amounts of product information and potential for immediate feedback and collaboration among product developers. 3D facilitates concurrent engineering, enabling different functions in the product lifecycle to take place at the same time, based on a common, single data representation and PLM process.

According to Kevin Marseilles, advanced manufacturing designer at Delphi, “The value of virtual 3D models is huge. You get so much more information from a 3D model. Physical properties help me perform virtual testing without ever creating a prototype, and I even provide CMM data points so my suppliers can check their parts against my digital model.”

Mike French, engineering manager at Aero Gear, thinks the biggest value of 3D comes from the associativity of their drawings with 3D models and the time saved in creation and management of op sheets. As Mike explains, “Using 2D CAD to do op sheets was very repetitive. Everything had to be manually edited. Now we can automatically populate a template with part information, and changes occur much easier and faster. We make them once and they flow through the process.”

There are business benefits too. Mr. French discovered that his customers like working in 3D, so they improved customer goodwill by implementing 3D CAD. And they expanded their business. Before they got into 3D they were mostly doing part manufacturing. Now they are taking on more design work.

3D software saves time. Numerous reports over the years have consistently shown 70% to 100% user productivity improvements in case studies. There is less duplication of work, for example, when drawings are generated from 3D models. Generation of an unlimited number of views is merely a matter of a math calculation, not person time. Creation of, say, a series of holes, is a simple task in a 3D CAD system, but requires creation of dozens of lines and circles in a 2D CAD system. 3D CAD reduces errors and improves quality. And 3D data models are more reusable, so designers can save time working from previous proven designs, rather than starting from scratch.

But if 3D is so powerful and valuable, why isn't everyone doing it?

One of the biggest obstacles to 3D adoption is that 3D software is still perceived to be expensive. Entry-level 3D CAD seats generally cost more than $3,500, plus options and support, which can be a lot for a small manufacturer. But looking at up-front cost is a mistake. The downstream benefits of 3D far outweigh the up-front cost. Marseilles agrees, “It’s a mistaken impression that 3D costs a lot. People are only looking at the up-front cost and just don’t understand the huge benefits in downstream productivity, design re-use, reductions in manual re-work and savings of manual drawing time, to mention just a few areas.”

Another reason 3D has not yet become ubiquitous is that it can be complex to use. The learning curve can be steep if the user interface isn’t well designed for the application. Randy Ochs, president at Actify (www.actify.com), believes this is one of the most important factors of a successful 3D application. To address this, they let users select their level of sophistication in their user interface, exposing or hiding certain functions depending on the sophistication of the user.

Mr. French agrees. As he explains, “Years ago, 3D was neat but too complicated and hard to use. The tools were cumbersome to use and the learning curve was straight up. Today's systems are night and day to old systems.”

There are also technical hurdles in moving to 3D. A lot of new designs originate from re-purposing of legacy CAD data. If you’ve been using CAD, your product data is already in some format you will need to be able to import into your 3D software. Standards such as IGES and STEP have emerged to help access legacy data. If complete 3D CAD models aren’t required, 3DXML (www.3ds.com/products-solutions/3d-for-all/3d-xml/overview) and the JT format from JT Open (www.jtopen.org) are popular formats for “light-weight” sharing of 3D models. If you and your customers or suppliers use a lot Dassault products, 3DXML may be a good choice, but JT appears to be more mature and robust, and claims a much larger consortium of over 100 companies.

And then there are the people challenges. Some people have been doing things “their way” for a long time and aren’t about to change easily. For these people, the benefits of any change need to be clearly exhibited, and it must make their lives and jobs easier than the status quo. Try as they may, a significant number of people just can't perceive information in three dimensions. A recent study by the Robert O’Connor Research Institute found that 20% to 25% of all people are unable to make sense of 3D objects.

Challenges aside, it's a sure bet 3D technology will add value to your business. So how can you benefit from it?

The first step is to define a product data management architecture and roadmap. Identify the primary creators and consumers of product data, and determine what data needs to move between them. Identify suitable data formats, based on current and intended data flows and applications. Consider your product lifecycle from concept design through production to retirement and consider what data will be required at each phase. This is best done concurrently as you are considering or implementing your PLM solution. It is relatively easy to model where you are, but much harder to look in the crystal ball and model where you want to be. But, as Alan Kay said, “The best way to predict the future is to invent it.”

Figure out where 3D fits and where it doesn’t for your operation and processes, today and in the future. 3D CAD maybe overkill for some users, but access to the 3D product data through a viewer can be a huge time-saver in purchasing and many other areas. Develop a roadmap for deploying 3D, based on how much of your business processes you plan to implement over time. For example, are you thinking of using 3D in sales or aftermarket service, technical documentation, spare part sales? Then align your 3D deployment roadmap with your IT roadmap to make sure the required resources will be available.

In an excellent article in May 2005 COE NewsNet (www.coe.org/newsnet/May05/index.cfm), Tim Ambridge from Bombardier outlines some of his basic lessons learned regarding introduction of new technology. Perhaps the most important is to focus on the 20% of the functionality that will provide 80% of the benefits. There are many bells and whistles in today’s 3D software market. As Mr. Ambridge says, “Do the greatest good for the greatest number of people.”

Then select and implement a pilot project. Choose a high-visibility, low risk project that you can accomplish in a finite time period and measure the results with and without 3D software. Analyze the business value and show management how 3D will help the whole organization.

And lastly, consider implementing a design methodology when you deploy 3D technology, much as you would implement Six-sigma or lean. By implementing their design methodologies, Delphi increased by 55% the time they saved using 3D CAD software, compared to using the same software on the same designs without a methodology.

Then stand back and watch how 3D software transforms your business. Deploying 3D is worth it. Your competition is probably already doing it. It can be challenging, but likely rewarded with increased productivity and profitability, or just survival – you choose.

David Prawel is founder and president at Longview Advisors Inc., a consulting firm specializing in 3D software technology in the manufacturing industry. www.longviewadvisors.com

Engineering Collaboration for the 21st Century
Bob McCandless, CEO, The McCandless Group

The technological, social, and global economic changes in the last few decades have made it imperative that engineering organizations redefine their product life cycle management solutions to include an ever increasing array of engineering collaboration tools. As we move forward into the first decade of the 21st century, the increasing pace of change in communication technology; advances in modeling, simulation, and manufacturing systems; and the requirement for knowledge-based engineering solutions will require a constant reevaluation of engineering collaboration tools. As we move farther into the future, the incorporation of artificial intelligence and expert systems into product development processes will drive a need for integration of new types of collaboration solutions; not only between individual engineers, but also between a wide array of roles and capabilities.

At the COE 2006 Show in April, my session will focus on the changes in engineering collaboration over the next few decades. In this session I will demonstrate detailed examples of methods for geographically dispersed teams to share and transform information into knowledge by combining and analyzing it in new ways. By combining engineering collaboration solutions with interactive, multisensory, immersive environments, experts in several engineering disciplines at widely distributed locations will be able to rapidly apply novel technologies, create better products in less time, and manage risks more effectively.

The use of multimedia and multisensory facilities will enable the engineering team of the future to explore, analyze, and understand complex phenomena remote from their everyday experience, and can help in displacing intuitive misconceptions with alternative, more accurate mental models. Advanced eLearning systems will provide the ability to reduce skill gaps. Visualization and simulation technology will enable the study of highly coupled multi-physics and multi-scale phenomena, such as those associated with computationally driven materials development, from the nanometers of quantum mechanics to the meters of operational engineering systems.

Future engineering teams will work in a highly heterogeneous environment (including different computing platforms, software systems, and multisensory immersive facilities) to organize and generate information and ideas. Simulating activities (such as data and information sharing), processes (such as generating and prioritizing alternative scenarios and decision making), and information flow during these interactions, can help in educating engineering teams, identifying guidelines for enhancing their performance, and improving decision making early on. It can also help in developing metrics to assess the effectiveness of the collaboration technologies deployed.

If you find this topic interesting, I hope you will be able to attend the session and contribute to the conversation. I will be setting aside plenty of time for Q&A at the end. I look forward to seeing you at the COE 2005 National Conference!

This session will overview the human side of Bombardier’s experience with the deployment of CATIA V5 and ENOVIA LCA on it’s new *CSeries passenger Jet. We will discuss how the new processes and the new tools were implemented to the various organizations and how they were received by the end users. The impact on all organizations within the company will also be discussed.

The session will offer experiences and insight into some of the issues you may come across trying to implement change in a large organization.

Implementing a PDM Installation and Migrating Data in a PLM Environment
Craig Miller, Purdue University, Professor
Nathan Hartman, Purdue University, Professor

The goal of this study was to gain an understanding of how companies in various industry sectors handled the migration of product data and the implementation of a PDM tool within the scope of a larger PLM environment. Ten companies were examined regarding their implementation process relative to PDM and PLM toolsets. Relevant sources from within each company were interviewed to gather information in a long interview format. Information was collected according to four different areas: PLM implementation timelines and mitigating factors, chosen PLM toolsets, data archival and migration strategies, and training.

In an effort to better understand the selection, implementation, and migration of PDM data and tool sets within a corporate environment, ten (10) companies were selected to participate in this project. They are all U.S. companies, although all of them have multiple divisions within this country and abroad. The participant companies represent the following product sectors: aircraft manufacturing (commercial, corporate, and military), aircraft engine manufacturing, heavy equipment, agricultural equipment, automotive manufacturing, automotive manufacturing suppliers (Tiers 1 and 2), and aerospace defense. These companies were selected based primarily on their ability to represent the collaborative environment necessary to derive full benefit from a PDM system. They were also selected based on their size (workgroups and numbers of corporate divisions) and their availability and willingness to participate.

The interview questions were developed according to the structure of PDM applications suggested by Cornelissen (1995): structure management, retrieval management, release management, change management and work flow management. In addition, the interview format also gave the researchers the opportunity to probe deeper into the reasoning behind the selection of a particular strategy or decision. The interview guide consisted on twelve questions that covered the following four major areas: PLM implementation timelines and mitigating factors, selected PLM toolsets, data archival and migration strategies, and training

In general, the selection, implementation, and configuration of PDM systems, as well as the migration of data between PDM systems, requires the development of a good process, the necessity for a champion at the upper levels of management, the need to organize and prepare your data before the migration starts, and the necessity to change corporate culture and the mindset of the users to accommodate the use of these new tools. When it comes to the actual migration process itself, the biggest point to be made is not to try everything at once. Form a set of “early adopters” who can help with the rollout, who can test various functions before recommending them to all users, and who can act as mentors within the different engineering groups once the toolset is released to everyone. In addition, it is important to communicate results of the migration to everyone involved, especially upper management. Finally, dedicated and consistent internal funding is critical to the success of this type of endeavor.

Deployment of CATIA/ENOVIA V5 and Their Impact on your Organization and the People in it!”
Colin Campbell, Bombardier Aerospace

Some of the areas to be discussed are;

• Where was Bombardier coming from and where do we want to go?
• Why did Bombardier Aerospace choose CATIA/ENOVIA V5?
• Where does it fit in our PLM strategy?
• How large was the impact on our people. How were processes changed, why were they changed, what was the impact on the people and their organization.
• What does no 2D release mean and how did we implement it? How does not having drawings affect everybody? Where does a parts list come from?
• How did we go about getting our organization to embrace the change?
• How do you change a herd of engineers’ intent on nothing but product design?
• What was the impact of the processes and tools changes?
• Where did we have problems?
• What lessons have we learned?
• What would we do differently?

Biography
25 years of aerospace design experience in both the rotary and fixed wing arenas. Started on the drafting board moved to CADAM onto CATIA. Have had the good fortune to work in a diverse number of design disciplines, from structural to systems design. Current responsibility is the manager of the process and tools group responsible for defining the Product Definition processes and tools that will be used to develop and analyze Bombardiers new C-Series passenger jet.

Implementing Product Life-Cycle Management - “The Good, the Bad, and the Legacy Data”
Richard H. Parr, Sikorsky Aircraft, LCA Project Lead

Purpose of this paper
Inherent in the high engineering design content of complex Aerospace and Defense (A&D) products, the processes involving all phases of product data management is mission critical. As A&D companies adopt virtual enterprise concepts, the information infrastructures must support inter-company exchange of product data along the supply chain and it’s customers. To meet these challenges of managing product data and their processes, Sikorsky Aircraft has been migrating to the next generation of Product Life-Cycle Management systems (PLM).

This paper will address how Sikorsky Aircraft is implementing Enovia’s Life-Cycle Application (LCA) software across their virtual enterprise. It will discuss the good, the bad and the legacy data aspects of this challenge.

Background – Transforming Sikorsky to accommodate a changing business environment
Managing product data in the changing aerospace defense environment requires a quantum leap past the early days of sepia drawings and dispersed print cribs. War stories about the effect of errors in mismatching revision levels of parts and documentation were common. Many aerospace company’s processes are drawing based, some pieces of the product never got into existing PDM systems, and they’re still kept on Mylar. They’ve created CATIA models against those drawings, but control effectivity via the Bill of Material, in their MRP system. Such was the environment many aerospace defense companies operated in during the past ten-fifteen years.

Sikorsky was operating in this environment until February of 2004, when the decision was made to implement a PLM system, replacing their existing Enovia PM Document Management system, their in-house designed DMU system, and their in-house highly customized Bill of Material system.

Figure 1 represents how data (models, bills of material and drawings) were controlled at that time, multiple systems, with multiple workflows, and data stored in three separate systems.

Figure 1 - Engineering Environment (2005)

Today’s existing PLM systems are no longer just for engineering; they are enterprise wide and manage the entire product life cycle. This requires the introduction of collaborative business processes to facilitate the everyday tasks performed by personnel in coordination with customers and suppliers. The fundamental need is to provide design-in-context, accurately, and efficiently throughout Integrated Product Teams and the supply network, while maintaining tight configuration control.
To meet this challenge Sikorsky is changing its business model, including sub-contracting out subassemblies of the aircraft for manufacture and sub-contracting detail design aspects to key contractor design facilities, while maintaining core functionality in-house. All assemblies being delivered back to Sikorsky in synch with our current build schedules. This scenario presents a whole series of challenges, for our team, which we will discuss later. The figure below shows the target architecture for collaboration.
It is always best to start with the good things when discussing implementations, so I will begin there.

The Good - Benefits
Engineering has seen a number of benefits from the use of LCA/V5 for the Conceptual Design project underway on the HLR program:

• Foremost has been improved collaboration and instant visibility of the work in progress on a given program. As one designer is modeling parts, other designers, structures engineers, design supervisors, mass properties engineers, manufacturing engineers, etc. have almost instant visibility to these parts. This has allowed them to identify necessary changes, such as aligning primary load paths, on the fly. It has also allowed for earlier input from manufacturing engineering, mass properties, etc. to help optimize designs at an early stage, when costs are minimal.
• The use of relational design has been a big time saver for modeling. In several sections of the aircraft, designers can swap out BCL models to change the shape of an area. They can now update affected parts and have them instantly morph to the new shape. They have also been able to quickly reposition groups of parts, such as a row of beams, using the local datum file feature, linked to the master datum file.
• Real-time updating of the LCA model repository (replacing the DMU) is a real benefit. No separate support group is needed to format and move models (typically a few days or weeks lag time) as on legacy V4 programs. This is the benefit of having one integrated Design-in-Context environment.
• Our CMO group initiates the skeleton model structure and as designers create details, they are forced to create part numbers for them and link them into the structure, thus building an accurate BOM to be used by both engineering and manufacturing from the get go.
• The ability to store multiple configurations, organized in study (non-BOM) GCOs and selectively load these options has been useful for trade study work.

There are always two sides to every story; in this case we have broken the other side into two separate pieces. First let’s discuss the challenges we need to overcome or adapt to.

The Bad - Challenges
As with any project there is always the bad side, or challenges as we like to call them, here are several examples of what we have encountered so far. Our current project is moving into detail design and therefore we are moving into a configured structure. We are also starting to move new assemblies from existing Helicopter variants into CATIA V5 and LCA, while keeping our legacy systems in synch. Let’s discuss several items:

• Supplier Collaboration presents numerous challenges with your internal networks;
• know about the real world,
• be able to handle that knowledge in a multi-disciplinary fashion,
• know how to structure and re-use knowledge, and
• allow automation (even support self-adaptation).

Response
Our team created a typical designer scenario and tested response times internally, at remote sites within Connecticut and at our remote design centers. We had demonstrations of response times with caching and replication. We also used a 3D XML model with our customer, which proved very promising. Tests showed our external networks were not adequate for heavy model usage; we are currently updating our network lines to several remote sites. Depending on size of models and total parts caching may be ultimately required.

Figure 2 - Supplier Collaboration

• Having a mixed environment of Catia V4 and V5 presents a DMU challenge to ongoing “Design-in-Context” efforts. As you transform your environment how does V4 designer take advantage of V5 data changes, and how can V5 designer take advantage of V4 changes?

Response
In our current DMU environment we had created a series of sub-directory files containing the V4 models for each of our current aircraft plus additional files for our key suppliers. Our team wrote a program to dumb down the V5 models into a “.sheet” and a “.model” file and populated the old V4 directories, while creating an index of file names. This allowed the V4 community to take advantage of changes done in V5 and continue to design in context with the newer designers as we migrate into our new environment. The V5 community could also read the directories and pull them into CATIA V5 for their work effort.

• Change Management ECO releases and propagations are online transactions only. There is no batch processes for these functions across multiple products and aircraft. We have large assembly’s which go down twenty-five plus levels with over 155,000 parts per aircraft.

Response
In our current MRP system we can not process online ECOs due to the shear volume of changes and our deep structures. We process 48 per day, 240 per week, with as many as 13,000 in WIP at one time. Changes are processed at night and our trickle down program applies effectivities to all affected Bills. We are currently testing cpu times with “a day in the life scenario” created by our designers and extrapolating those times across a user population slated to be over 2,000 users when we are completed. We are also testing an ECO across three distinct models to examine response times at various levels. Our first actual test of LCA ECO release across three test models took two and a half minutes. Testing will continue, more models deep levels.

• Working with Configured Products requires too many steps for the Design community. Application of effectivity requires too many manual interventions, between actions and modifs.

Response
Our current system has had customizations applied to eliminate numerous screens allowing just one entry of the effectivity. We are contemplating similar customizations to LCA for our design community.

• Working with Flat BOM. Many of our legacy BOMs have twenty plus levels within them. Dassault advised us to flatten them for improved processing. May be very difficult for legacy aircraft.

Response
For our first new program called “HLR” attempted to go with a totally flat BOM. We used GCOs initially for ease of building our conceptual model. However as we got closer to detail design phase we learned we must have a configured BOM, and working with a totally flat configure BOM did not work. We reconfigured into a BOM with about 7 levels, at its deepest point, not flat but flatter than any other existing program.

Legacy Data - Hurdles
As you saw from are original slide our environment has data in multiple systems and to some extent information is duplicated. Information is also susceptible to errors as the data is spread across those same systems. Trying to merge these systems in a timely manner presents a huge challenge to our team.

• Migration tools from Dassault were not conducive to our large databases, with over 2,500 helicopters. The main tool was STEP, although very precise, it would take too long. In our case we were talking over one year to migrate the data, and with number of changes we processed it might never end.

Work Around
Our team, working with subcontractor, developed a direct database load program.

• Initial test of 155,000 parts for one legacy Helicopter program took 25 hours, not good. More expertise brought to bear.
• Second test took 26 minutes, now were getting there.
• Dassault “Checker” tool validated our data load-found errors
• By third load we had a clean database.
• Rigorous testing now in plans, recently loaded 3 helicopters and models in less than 60 minutes. Processed transactions against all three and performed an ECO across all three models (2 ½ mins)!
• We are now testing scripts against each model and across all three.

• Legacy MRP system was a highly customized system evolving during the past fifteen years and contained many functions particular to our day to day operations.

Work Around
Our team is examining many alternatives to try and avoid making all the same customizations to LCA. This could be elimination of our EBOM and MBOM into a common BOM for processing by all systems. Alternatively, LCA could process our EBOM and SAP our MBOM. We are also investigating the use of an "Effectivity Engine" in between to pick up functionality such as "Breakpoints".

What would we have done differently?

• Kept customizations to a minimum.
• Try not to mimic our legacy systems, but address the business problem.
• Work closer with software developer, not the system implementer.
• Worked “day-in-the-life” scenario into initial Proof of Concept
• Made more plant visits