- Welcome Page 2 - Update Page 3 - News Page 4 - Brief lines | Welcome to this issue of 2PLM, an e-zine distributed every two weeks on a free subscription basis. This week Vijay Srinivasan of IBM writes about "Product Data Exchange: Geometry and Beyond". * Product Data Exchange: Geometry and Beyond * Recently, there has been a flurry of announcements and discussion around the topic of product data exchange and the role of open standards in such an exchange. It is time to take a deep breath and reflect on what we at IBM are trying to achieve and how the industry at large is approaching these issues. IBM is embracing open standards as part of our overall business strategy. Industry sees value in this approach and has already made some sensible grouping of product data for standardization. For discrete, engineered products, it is useful to think about the product data consisting of two types of information: one is geometry and related attributes, and the other is bill of material (commonly known by the acronym BOM) and related attributes. In the past, the reason for this split was to facilitate the understanding of various industrial activities clustered around these two types of information. Geometry and related attributes resided in the domain of engineers. In fact, these were left as "engineering objects" to be dealt with using geometric and graphics applications. On the other hand, BOM and related attributes were considered "business objects" because they formed the basic units for operational management, including inventory, supply-chain, and sales. Part numbers dominate much of the discussion around BOM and related services. It is not surprising that CAD/CAM/CAE applications deal with a lot of geometric (and graphics) data, whereas Enterprise Resource Planning (ERP), Supply Chain Management (SCM), and Customer Relationship Management (CRM) applications treat BOM as the basic unit of information. However, there is clearly a link between geometry and BOM, which is highlighted in Product Data Management (PDM) applications which address both, - and this should not be ignored. While we are discussing geometry, it is worthwhile to distinguish between geometric data, which demands a high degree of fidelity, and graphics applications, where we are allowed to "cheat". In fact, geometric information exchanged in the form of dimensioning and tolerancing are legally binding specifications - just follow the number of court cases that involve disputes between parties on such specifications and the number of consultants who make their living as expert witnesses. The game is different in graphics - here we are mainly concerned with the human eye visualizing the product geometry and how much we can get away with cheating the eye. This explains why an open standard such as ISO/STEP (STandard for Exchange of Product data model) focuses on computer-interpretable geometry instead of graphics, and this allows others to define various types of graphics standards. Some of the graphics standards make up for their deficiency by capturing high fidelity information, such as dimensioning and tolerancing, using graphical and textual annotations that are left to be interpreted by humans. At the basic level, the BOM is an (unordered) list of a complete set of parts in a product, with no redundancy. This seemingly simple definition might give the false impression that it is easy to manage and exchange such data. While it is relatively easy to define such data types, those who have dealt with BOM management know how difficult it is to ensure the correctness, completeness, and non-redundancy in such data. It is no wonder that BOM and related attributes are the focus of attention in many e-business applications and standardization in industry. In reality, to be honest, one has to deal with structured BOM, which maps to the assembly structure of the product with links to part geometry and spatial relationship between parts in that assembly. This is a challenging task, but it has been successfully tried in a few industrial pilot projects using open standards. One such pilot project was conducted by an Aerospace Engine Alliance, consisting of major aircraft manufacturers and engine suppliers. In this project, Airbus in France and Pratt and Whitney in the US exchanged very large assembly models of aircraft engines - with detailed geometry and product structure - using open standards (in this case, STEP.) These two companies have been using different CAD and PDM systems, and exchange of the product information using a neutral format was the only viable, long-term option they had. This was not just an engineering data exchange problem, because it also impacted the business process used between a prime and a lead supplier. The result was so successful that they have been using it in production. Again, in this example, it can be seen how the traditional simple split between "engineering" and "business" objects and activities is breaking down. As we move beyond PDM to PLM, previously separate activities and data are being joined up across the product lifecycle. Historically, the aerospace industry has led the product data standardization efforts. But the results are trickling down to other industries. A few years ago the Automotive Industry Action Group (AIAG) conducted a pilot study with funding from the National Institute of Standards and Technology (NIST). This study involved XML-based standards for BOM along with STEP-based standards for geometric models. They concluded that standards-based collaboration can work in a global, distributed, and heterogeneous design environment; this established the technical feasibility of such an approach. They also concluded that such internet-based technology solutions are inexpensive, readily available, and easy to deploy in the supply chain - thereby proving the business case. Stay tuned for more success stories. For questions and comments, please mail me. Vijay Srinivasan, Ph.D., is with Research, Standards, and Academic Programs, Worldwide, IBM Product Lifecycle Management.  * PLM articles * Thanks to readers for informing us of these articles about PLM in other publications:
* Join the Product Lifecycle Management Interest Group * The PLMIG is an Interest Group for organizations wanting to proactively improve their PLM performance. The scope of PLM is wide and its implementation lengthy and complex. By working together in the Interest Group, organizations can achieve expected PLM benefits quicker. The PLMIG started operations in January 2004, and by December 2004 had about 50 member organizations from 15 countries. Member organizations include engineering and manufacturing companies, PLM application system vendors, integrators, consultants, universities and research organizations. The PLMIG is funded partly by member subscription fees, partly through its activities (e.g. publications and events). Publications include the regular monthly PLM Journal as well as reports resulting from activities. In 2004, PLMIG activities included 5 one-day meetings, the preparation of 2 proposals for collaborative PLM research projects, and a three-day PLM Benchmarking Workshop leading to the publication of the PLMIG's PLM Benchmarking Handbook. There are 2 classes of PLMIG Membership, Individual Membership and Corporate Membership. Individual Membership allows one named individual from an organization to participate in PLMIG activities and receive the PLM Journal. Corporate Membership allows five people from an organization to participate in PLMIG activities and receive the PLM Journal. The strategic direction of the PLMIG is set annually by the Corporate Members, meeting as a Steering Committee and considering proposals from all members. The direction for 2005 was set in December 2004 by the organizations that were then Corporate Members, two industrial companies and two PLM application system vendors. In 2005, the PLMIG will be working towards identifying PLM Best Practices for its members, using an approach based on benchmarking PLM performance. Benchmarking is considered as essential for achieving a clear, factual view of an organization's PLM activity, from which improvements can be made. The benefits of PLMIG membership are different for different types of organizations. For manufacturing companies, PLMIG membership provides an opportunity to collaboratively work with other companies to solve common problems. For universities and research organizations, participation in the Research Special Interest Group (RSIG) provides a platform for collaboratively preparing research proposals. The PLMIG includes a large number of organizations focused on improving their PLM performance, and willing to collaborate with others. If you want to improve PLM performance, become a PLMIG member now. Contact Roger Tempest for membership details. * PLM Benchmarking Handbook * The PLMIG announced that further to its November PLM Benchmarking Workshop which brought together participants from many countries, companies and industries (including electronics, medical and automotive OEMs and suppliers), the PLM Benchmarking Handbook is now available. It can be used to conduct PLM benchmarks, internally or with external partners, in any industry, at any desired level. Details |
