Best Practices in CAD Data Exchange

Background
Manufacturers face relentless pressure to develop new products and bring them to market faster than ever before. Twenty to thirty percent of all sales by industrial companies come from products that have been on the market for less than 5 years. Competitive pressures drive innovation and demand quality, while increasingly complex products have led most companies to depend on global business partners for component parts.

To win, companies must integrate their customers and partners directly into their product development operations. Success depends on how well business partners share information and react to changes in design and process, regardless of their geographic location. Much of this success relies on fast, efficient sharing of product data.

The challenges and difficulties of CAD data exchange are well known and publicized. Studies have suggested that the cost of problems in sharing engineering data to the automobile industry alone could be $10 Billion per year. Few single issues present a more formidable impediment to accelerating product time-to-market. Large companies have implemented significant procedures, tools and infrastructures to try to alleviate the huge impact of these problems, but generally push the problem to their suppliers. Suppliers, generally unable to afford sophisticated tools, have little more than brute force methods to address these problems, as pressure builds to further accelerate product development, improve quality and cut cost.

Interoperability best practices are poorly understood. This paper presents the results of research that investigated interoperability best practices at 50 small to medium discrete manufacturers, to better understand what works and what doesn’t work for manufacturers who build products on a local or global level. Our highly qualitative research discovered interoperability best practices from a broad range of discrete manufacturing industries and company sizes, and representing a variety of CAD platforms.

Our study confirmed that the costs and frustration resulting from poor product data interoperability, and the negative impact on manufacturing productivity are indeed very, very large. Although the most costly and visible problem derives from poor product data quality resulting from different CAD systems and formats, significant costs are also associated with lack of data communication tools to make the processes easier and smoother, and lack of management support to address and manage these problems.

Common causes of poor data quality also cited in our research include poor CAD model creation, resulting from a lack of training or sloppiness or both, and mathematical inconsistencies naturally inherent in CAD systems that use different modeling techniques and algorithms. For example, edges and faces that have small gaps between them, “sliver” faces (very thin aspect ratio and generally very hard to detect), and duplicate vertices and edges, and faces with inverse orientation. These errors often result in an engineer re-creating the offending part of the model, resulting in the largest single direct cost factor in data exchange.

Dassault Systèmes (Spatial) and other vendors have made recent advances in “healing” technology which shows interesting promise in automating the repair of many of these problems, but it is not yet a perfect nor fully-automated solution, and may never be.

Further magnifying the cost of poor quality data translation is the fact that many error conditions go undetected until later in the design process, where they create additional, potentially much more expensive problems. For example, an error in a CAD file may not appear to CAD users, but may appear during finite element analysis, or when engineers are creating physical prototypes, or even further downstream (and more costly!) when manufacturers are developing tooling. Different applications put model data to different uses, each potentially exposing different effects of poor data quality, and all causing model re-work and project delay. The further downstream data quality problems are detected, the more costly they are, both directly and indirectly. It is not uncommon for companies to report that downstream functions, such as rapid prototyping, finite element analysis, or CNC programming, spent as much as 50 percent of their time working with CAD data files that exhibited quality problems of some sort that impeded their efficient use in downstream applications.

A survey reported in Interoperability Cost Analysis in the Automobile Industry, published by the Research Triangle Institute in March, 1999, revealed that in about 51 percent of the jobs, the CAD data had to be repaired. The job shop had to completely recreate CAD data in an additional 25 percent of the cases. In about 15 percent of all cases, these errors were not discovered until after the part tooling had already been cut, resulting in significant cost and delay because the company had to scrap and re-cut the parts.

The largest manufacturers are accustomed to dictating the CAD systems used by their suppliers and have tried for many years to enforce this requirement. This may appear to be the simplest way to minimize interoperability problems, but unfortunately it’s not that easy. In fact, this may be the most expensive method when all direct and indirect costs are considered. Supplier costs have to be absorbed either in quality, time or price. It is common for suppliers to be forced to have one of each CAD system, and fully trained staff to use it. At the end of the day, this cost comes back to the OEM indirectly as higher cost to their suppliers. Other factors also contribute to an apparent trend away from this approach to interoperability. No single CAD system can be the best for all, so sacrifices have to be made. These sacrifices translate into indirect, often intangible, cost, such as in lost opportunity for innovation or time to market. A very large, global manufacturer we interviewed reported no cost advantage to having everyone using the same CAD system, so they have ceased this practice, making these decisions on a program basis.

Among the neutral formats, such as STEP and IGES, IGES is by far the most popular industry standard neutral file format among the professionals we interviewed. We found many users quite satisfied with the results from IGES file transfer in many 3D situations. One of the disadvantages of neutral format translation is that interoperability involves many organizations, such as engineering, purchasing and marketing, and product data includes a very wide range of data types, not just CAD data. IGES is very limited in its ability to deal with different data types and applications. STEP was designed for this reason - to be extensible into many more functional and business requirements. However, its huge scope (and high cost to develop and support) appears to be its downfall.

Although very popular a few years ago, STEP did not appear in our research to be widely used, except among the largest aerospace companies and in some specific applications where only the simplest STEP files are required (e.g. with no features or history). The main reasons expressed for the lack of STEP usage were high cost of the translators from the CAD vendors, and the fact that the standards bodies generally move too slowly, so the formats never seem to evolve quickly enough. One large worldwide manufacturer we interviewed has ceased all participation in any of the standards activities due to low rate of visible results & low ROI. Another told us they thought “STEP has now become the white elephant”.

Most data exchange discussions center on CAD data. However, a huge amount of non-CAD data is also shared on a daily basis. We found that the most preferred method for sharing these data is to email Adobe PDF files. Probably not a surprise – it’s inexpensive to create them (using Adobe Acrobat) and free to read them (with Adobe’s PDF viewer).

Direct translators are available for a wide variety of CAD formats. These can be purchased from the vendors of the CAD systems or from independent suppliers. In general, companies we interviewed noted that after numerous years of testing and use, they felt that translators from the independent suppliers were better and less expensive than those from the CAD vendors. Commercial translators are considered expensive, but a thorough analysis often reveals acceptable ROI. We found that some popular commercial translation software worked quite well, especially for common problems with CATIA file translation. For example, Theorem Solutions (www.theorem.co.uk) was most often mentioned and respected, but there are many good vendors to look at. There is no substitute for good testing, including a wide range of data typical of your business, and “round trip” testing, to and from another format , producing the same results as verified by good quality checking tools, such as I/Check from Incat (www.incat.com).

A tremendous amount of engineering resource is spent performing manual tasks associated with sending and receiving data, confirming receipt, tracking contact information, etc.–tasks that could easily be automated. A large aircraft manufacturer we interviewed reported that up to 25% of an engineer’s time can be spent solving data exchange problems, in larger projects. This time is mostly spent transmitting and retransmitting data and re-creating faulty models. At their quoted rate of thousands of models exchanged in a typical month, this adds a huge cost to the program. So it’s easy to see why the cost of interoperability is so high. If problems exist in only a small percentage of all product data files exchanged, and these problems cause just an hour of rework or lost progress. This represents a huge detriment to productivity.

Email is far and away the most common form of communication between parties. Email is used throughout the business process, internally and externally. Most communications between engineers and purchasing or suppliers is done through email. “It’s not efficient, and all the back and forth costs a lot of time, but it works with relatively little hassle”, was often heard in our research.

The second most common form of communication is to put a CD in the mail–low tech but it gets the data there, usually. When files reach a size of about 10Mb (compressed), they are generally put on an ftp site, or put on CD and sent through the mail.

A significant and virtually unaddressed aspect of interoperability is the underlying infrastructure to support effective data communication. Currently, it is up to the engineers to know who should get file(s) at each particular partner company, email addresses and contact information, required formats and versions, log and archive transmission events, notify receipt of data, know who's authorized to get data and not, etc. We have found this to be the second largest source of cost and wasted time associated with interoperability problems. A large worldwide manufacturer we interviewed felt that this is the best area to invest in for the biggest ROI–minimizing and streamlining cycles between relatively high paid professionals.

One highly regarded data communication solution is DDX from ProCAEss GmbH (www.procaess.com). Another new product that addresses these issues is called Fusion-DX from Datranet Ltd. (www.datranet.com) and an often-mentioned solution for the automotive market is the AutoWeb (www.autoweb.com).

There are many reasons senior management doesn’t “get it”, but the most common is that interoperability is generally perceived to be a technical issue, to be solved by technical people. Interoperability problems are often associated with technical issues like model format problems and other things that senior execs don’t understand, nor want to understand. Thus interoperability solutions never seem to make the short list in the budget.

Some companies understand the cost of poor interoperability business practices, but we found no one who attempts to measure it. A few companies appear to recognize the significant opportunity cost associated with poor interoperability practices, and the productivity value of efficient supplier interaction and communication, but for the most part, these costs are passed on to the program or project.

There are many methods of exchanging product data, and each has its advantages and disadvantages. There are many factors involved in selecting the right method for your company, for example, the CAD systems used at each end of a transaction, the number of data files sent, the size of the typical file, the number of suppliers, and many more. The best system will result from a careful evaluation of the issues discussed below, and a clear management dedication to attaching visible importance on being good at interoperability.

Perfect interoperability should never be the goal for all transactions–just sufficient to do the job, and scalable enough to adapt to new projects where the requirements may be more rigorous.

CAD training is a must. We found that there is no substitute for investments in training for CAD users. Interoperability training and consulting is valuable, especially when new problems arise or new projects involve new types of data or other requirements. Engineering managers we interviewed told us that training their CAD users can reduce by 15% to 25% the interoperability problems they experience.

The most dedicated people currently working on interoperability problems in your company are probably self-designated champions of the cause. The more support your internal champions get, the more effective they can be. Management visibility and recognition of the magnitude of the problem can help support the interoperability person or team, at the highest possible level, and help reduce your interoperability costs.

Consider implementing an interoperability team–a cross-functional team that should be managed as a corporate function. Empower them to develop and implement policies and processes for interoperability. Good methodology can prove much more valuable than the latest technology. Capturing and communicating the insights of your most experienced people, perhaps through a “Portal” or message board, will help reduce interoperability problems and costs, since success rates appear to improve dramatically between experienced data exchangers.

Your business process analysis should include consideration of who will take responsibility for conversions and the final digital product delivered. In general practice we found the supplier will be expected to deliver 100% correct, fully inspected data, on time. CAD quality checking tools can help in this area, but these tools they are not sufficient, and they have not yet been certified by standards bodies like SASIG or AIAG. So be careful assuming too much liability protection from quality checking tools.

A dedicated data transfer person is a wasted engineering resource. Invest in good translation and communication tools and develop win-win relationships with their vendors. Let your engineers do engineering and the purchasers do purchasing. While it is dangerous to generalize too much, we recommend that if your engineering team transmits more than about 100 files in a typical month to and from a few dozen suppliers, you may be able to justify the cost of a communication solution with the time your engineers will save in manual file sharing tasks.

Competitiveness in the global manufacturing industries has amplified the importance of interoperability solutions and the urgency of competence in sharing huge amounts of product data. As trade export imbalances make front-page news, global manufacturers have looked at many forms of business practice optimization, such as lean manufacturing, six-sigma, concurrent engineering, and the like. Interoperability problems are highly costly, but often underestimated and generally not noticed by management, as they secretly undermine productivity and time to market. Problems in interoperability affect business between companies and between different groups in the same company. It should be respected as core business process in any manufacturing operation and tantamount to success and competitiveness for any manufacturing company.

Whirlpool Adopts Digital Manufacturing Solutions From DELMIA Corp.

World's leading manufacturer of major home appliances links design and manufacturing data for production of new microwave ovens

DELMIA Corp., a Dassault Systèmes announced that Whirlpool Corporation is embarking upon digital manufacturing technology. Whirlpool is achieving this goal thanks to DELMIA Process Engineer and V5 DPM Assembly solutions, linked by the DELMIA Manufacturing Hub.

Anders Claesson, production manager at Whirlpool in Norrkoping, Sweden, explained, “With three new microwave platforms about to be introduced and no integration between our materials and planning system, we recognized it was a good time to move to a digital environment, allowing the sharing and re-use of data across disciplines.

“In the past, our simple database did not connect CAD data with assembly times or cost of materials and there was no link between the ordering of tools or materials. After studying the marketplace and learning that DELMIA’s Manufacturing Hub could be integrated with our current PDM system, ProductCenter, we realized that product and manufacturing data could be transparently shared. In addition, we will improve our concurrent engineering using Process Engineer and V5 DPM Assembly for process verification and documentation,” continued Claesson.

Before adopting the software, Whirlpool ran a pilot program of the DELMIA tools. DELMIA applications engineers were stationed on-site at Whirlpool, helping to implement DELMIA and the work flow around the software, as well as integration with ProductCenter. All the data entry is now complete and Whirlpool’s designers are exploring their virtual prototypes in 3D.

The first family of microwaves has been completed and engineers are gearing up for the next product line. In addition to designing the microwaves in 3D, Whirlpool will rely heavily upon the DELMIA Manufacturing Hub, a data repository that stores both historic and current product, process and resource information. Engineers can continuously update and share the most current data to better manage all processes and equipment orders. Manufacturing processes that can be created and evaluated include time analyses, rough balances, ramp-up scenarios and capacity analyses. The various scenarios are stored in the Hub, allowing engineers quick access to information for reuse and to support decision making.

It is conceived that, in the future, Whirlpool’s Engineering Department will be able to provide CAD models and the engineering bill of materials (EBOM) to develop the necessary assembly processes and manufacturing bill of materials (EBOM) in one seamless and interactive process. The structuring of data analyses and calculations will be handled by V5 DPM and Process Engineering applications.

“Ultimately, we anticipate that DELMIA technology will accelerate time-to-market through faster product and process verification and validation,” said Claesson. “Manual data transfers should become a thing of the past and our assembly line workers may even complete their training in a DELMIA environment.”

About Whirlpool Corporation
Whirlpool Corporation is the world's leading manufacturer and marketer of major home appliances. The company manufactures in 13 countries and markets products in more than 170 countries under major brand names such as Whirlpool, KitchenAid, Roper, Estate, Bauknecht, Ignis, Laden, Brastemp and Consul. Whirlpool Corporation is also the principal supplier to Sears, Roebuck and Co. of many major home appliances marketed under the Kenmore brand name. Information about Whirlpool is available at http://www.whirlpool.com

SMARTEAM Simplifies Bidding for Mid-Sized Manufacturers
Shelli Zargary, Marketing Projects Manager, SMARTEAM Corp.

SMARTEAM provides a full best-practice-based bidding capability to industries in general and electronics manufacturers and industrial OEMs in particular, helping meet the challenge to deliver built-to-order and variant product solutions while putting out more accurate bids faster.

Perhaps no PLM-delivered capability drives faster and higher returns than the bidding process, which enables companies to more quickly process RFQs and produce more competitive, profitable quotes on time, for improved bid win ratios and increased revenues.

Bid preparation is a key bottleneck to profit. SMARTEAM provides key capabilities for optimizing profits through better bidding

SMARTEAM leverages a variety of its best practices - including Product Knowledge Reuse, Engineering Change Management, Collaborative Engineering, and Bidding & RFQ Management -- to fully manage and automate the bidding process. The SMARTEAM bidding process drives bidding from the engineer’s receiving of the bid request, through development of a 3D model, across all activities related to processing the bid, including the leveraging of subject matter experts within the enterprise and across the product value chain and to the submitting of the bid. SMARTEAM gives companies further advantage by enabling them to reuse and compare previous bids.

Using SMARTEAM, manufacturers following a simple, standardized bid execution path:

1.	Initiate Bidding Workflow - Engineering Change Management best practice
2.	Define and Cost Initial Proposal - Product Knowledge Reuse best practice
3.	Derive “As Quoted” project from "As Designed" template - Collaborative Engineering best practice
4.	Verify Component Costing - Supplier Relations Management best practice
5.	Proposal Management, with Web-Based Customer Collaboration and Bidding Workflow Management - Bidding and RFP Management best practice

These SMARTEAM best practices streamline bidding through pre-defined, standard processes for product configuration requirements analyses, maximum design reuse and accurate product costing. SMARTEAM further facilitates the submission of 3D "As Quoted" product models, along with all relevant specifications, drawings and other technical documentation, further enhancing the strength of the proposal.

SMARTEAM delivers its bidding best practice through demos, presentations, templates with automated bid creation workflows, data models, methodology guides, and supporting white papers, enabling companies to quickly implement the bidding capabilities with minimal customization. Solution demonstrations include:

	Mold Bidding Demo, with a scenario for a web-based customer request analysis and input from product area knowledge communities
	Bidding Process for BTO Demo, with a scenario for the Machinery & Mobile Equipment sector that covers workflow management of the customer's requirements, adaptation of the product definition to customer requirements, evaluation of part reuse possibilities, and generation of a quotation

Leading manufacturers* using SMARTEAM to improve their bidding processes and increase revenues include:

	AMC Centurion, Swedish manufacturer of antennas for mobile phones, relies on SMARTEAM for fast access to product information and accelerated supply chain responsiveness, achieving reduced RFQ turnaround time. SMARTEAM has improved the company's capacity to be responsive to customers, helping win more business.
	Krebs Engineers, U.S.-based supplier of hydrocyclones for a variety of industrial applications is expanding its use of SMARTEAM to improve responsiveness to customers via web-based bid processing, targeting 24-48 hour RFQ turnaround and the delivery of comprehensive proposals complete with designs, documentation and even case studies.
	Esmena, international supplier of often-customized warehouse storage systems, uses SMARTEAM to more quickly and accurately generate bids. By avoiding errors in costing, delays and slow-moving schedules, Esmena wins more contracts and completes projects more efficiently and profitably.