PDM for the enterprise

As manufacturers redefine themselves, software vendors are developing versatile tools that enable customers to re-engineer business processes locally and globally.

OVER THE PAST FEW years, product-data-management PDM) software developers have raced to develop World Wide Web-based interfaces to their products. Such interfaces enable staff outside the engineering department to gain access to limited views of product information early enough to have a positive impact on a new product's cost, quality, or time to market. Now, companies are focusing on building PDM systems from the ground up using Internet and related technologies. The goal is to provide a much richer feature set and knowledge base to an even larger community of users, whether they work in a manufacturers engineering center, at an overseas subsidiary, or at a supplier or other business partner.

Perhaps the most notable example of this trend is Windchill from Parametric Technology Corp. (PTC) in Waltham, Mass. In this exclusive interview with Mechanical Engineering, James Heppelmann, senior vice president responsible for Windchilll and PTC's overall information technology strategy, explains what has changed in manufacturers' competitive environment and how the new enterprise-wide PDM systems fit in this new landscape.

Heppelmann was co-founder and chief technical officer of Windchill Technology, a Minneapolis-based developer of Internet-, World Wide Web, and Java-based information-management software. The company was formed in 1996 and acquired by PTC last January Previously, Heppelmann was chief technical officer at Metaphase Technology Inc., a division of Structural Dynamics Research Corp. in Milford, Ohio.

Mechanical Engineering: How has PDM evolved over the years? Describe the PDM arena enterprise POM has entered.

James Heppelmann: PDM started as an intelligent file manager add-on for computer-aided-design and computer-aided-manufacturing (CAD/CAM) systems. Today, the focus is much more on information systems and bill-of-materials applications. An enterprise PDM system is the main central repository for all that there is to know about the product definition and all the many iterations of that definition. PDM is growing increasingly sophisticated. Take product configuration, for example: If the PDM system knew the features and options that a product could have, manufacturers could generate bills of material for product instances that have not yet been created.

ME: Does enterprise PDM redefine the concept of the product model?

JH: It's certainly become a broader concept. For every part in a manufactured product, there are many important documents-according to one study I've seen, there's an average of 15 documents per product-and only one of those documents is the product drawing or model. The broad view of PDM now is that while geometry continues to be important, there are 14 other definitions that are important, too. Such additional documents might include purchase orders, fabrication plans, or, perhaps, safety analyses.

ME: How does the redefinition of PDM affect the way manufacturing companies view the product information generated by an engineering department?

JH: If the engineers are the only community creating and looking at CAD models, then, quite frankly, you're not getting a full return on your CAD investment. But if you can find a way to leverage this information in some format that is easy to use, easy to understand, and easy to access for a much broader community of people, then that information is far more valuable.

Look at who is involved in the life cycle of a sophisticated product. Marketing people are deciding what the product should be, which versions of it are salable, and which ones will never be successful. Engineers are designing the product. Manufacturing people are fabricating it. Service people are out in the field performing maintenance and repair. There are many people who need to tap into a repository of product information above and beyond the engineer who created a geometrical representation of the part in a CAD file. So it's all about leveraging information as opposed to simply managing it.

That's the essence of PTC's strategy with Windchill. It will allow manufacturers to leverage the product model across the enterprise, to make the CAD model far more valuable by relating it to the big picture.

ME: Has the definition of enterprise changed?

JH: From a PDM perspective, I see the enterprise as including everyone who has a vested interest in the product definition. But it tends to go beyond the boundary of the company, because if you look at any major automotive or aerospace product, it's not built by a company; rather, it's built by a whole series of companies across a supply chain. In fact, in a lot of automotive companies, their core competencies are sheet-metal bending and final product assembly, and nearly everything else is outsourced. So for an automotive manufacturer, the enterprise involves a tremendous number of suppliers and constituencies not only within the company proper, but also across the supply chain.

This new, virtual enterprise is an exciting place where Windchill has applicability. Windchill takes PDM technology and injects it into a World Wide Web/intranet architecture, making product data immediately available to a broad community of people who don't need any special access tools above and beyond a Web browser.

ME: What are the market and technological forces that are driving the development of Windchill?

JH: The most significant change is the advent and the widespread adoption of intranet and Web technologies. In the past, companies had systems that were difficult to install, implement, and maintain across the broad enterprise-user community-there were simply too many computers to visit and too many people to train. Web technologies give us a way to take advantage of this incredible network. It's everywhere, it's relatively inexpensive, and it's highly usable above and beyond PDM.

What we're trying to do with Windchill is to add value to that trend, so a company that's already investing in intranet or extranet technology can use Windchill to serve up product information through that same information infrastructure.

Windchill is doing a good job of leveraging this technology. It's a much easier system to implement, and the ongoing cost of ownership is significantly less. In effect, you're getting the centralized management benefits of a mainframe with the runtime, distributed computing characteristics of a client/server system. It's intended to provide the best of both worlds.

ME: What is the problem that Windchill was invented to solve? JH: Flexibility is a key issue here. Manufacturing companies generally start by identifying the process by which they produce products-the "as-is" process. They know there's a better way to do it going forward-the "to-be" process. Companies want to put the "to-be" process in place because they believe it will decrease time to market, increase quality, decrease cost, and do all those other good things. Then they ask, What kinds of technologies can we procure that will enable us to realize the "to-be" process? They go out and procure technology, they implement the technology, and they move to the "to-be" process. And then they iterate through that whole cycle. Suddenly they're right where they were two years ago, needing to institute process improvements.

The problem is this: With a lot of information systems, a feature that starts out as an enabler soon becomes a constraint. It's like a gun with one bullet in it. You shoot the bullet, and it's gone. It helped you implement yesterday's process, but now it constrains you from moving to tomorrow's process.

Manufacturers are constantly redefining the boundaries of the virtual enterprise, redefining who their partners and suppliers are. They're buying companies, they're spinning off companies, they're reorganizing, they're starting new projects, they're retiring old projects-it's a very dynamic environment. An information system that freezes in place one snapshot of a manufacturer's processes at a given point in time becomes a real problem. In fact, it soon becomes what's known as a legacy system.

This development raises two questions: How can the enterprise use software technology to enable today's process, and when the time comes to redefine the process, is it fairly convenient, or at least possible, to redefine the technology? We think Windchill has some significant technical advantages in both of these areas.

ME: So your goal with Windchill is to enable a company to redefine its PDM technology as needed?

JH: At the root of this discussion is a top-down versus a bottom-up approach in implementation. A top-down approach demands one process that everybody in the company is going to use. It may turn out that your company makes cars and trucks, and it has a defense division that makes tanks. Given the different requirements in these different areas, it is not convenient to standardize on one common process for everybody. Doing that takes forever, and inevitably it yields a suboptimal solution for everybody. With Windchill, we're looking at a less bureaucratic, more autonomous model of cooperation, where people can implement what works for them but still cooperate with the next person even if his or her process isn't identical. The way I like to put it is, "optimize locally, integrate globally."

ME: What is the relationship between Windchill and PTC'S existing PDM offering Pro/INTRALINK?

JH: Windchill is actually more focused on the community outside of engineering. Many people would agree that for every user of the product definition within engineering, there are probably 10 users beyond engineering. An engineer who spends most of his day using Pro/ ENGINEER may rely a lot on Pro/INTRALINK because his focus is in that environment, whereas the enterprise user working with Windchill needs a much broader view, but one that is less deep. The enterprise user doesn't need to go deeply into a Pro/ENGINEER product model, but may need to understand how the product model and some purchased components fit together in this configuration of the product. Such a user is not going to be modeling any components but will want to know which components were used, who they were procured for, at what cost, and so on.

ME: So engineers who work in the enterprise PDM environment can know that their information will be easily accessible to other people.

JH: Absolutely. They can know that the people who need to get at the information they worked hard to develop now have a means to access it. CAD systems are sophisticated. Asking a procurement person to fire up her favorite CAD tool to look at the model just isn't going to fly. We need to get this information out to people in a format that is lighter in weight. We need a PDM system that says, Fire up your Web browser, go to this page, and look at the configuration, and if you find a part that interests you, select it and see a quick image in a very simple viewer, rather than use a CAD tool.

ME: What is the software tool set inside Windchill that accomplishes enterprise PDM?

JH: The first release of Windchill contains four module packages, the Windchill Foundation, the Windchill Information Modeler, Windchill Document Management, and Windchill Configuration Management. The Windchill Foundation is a general-purpose, reusable framework technology designed to host Web-based information systems. Windchill Information Modeler contains integrated development tools that help you use Foundation to model actual applications and create solutions.

Document Manager and Configuration Manager are examples of how we ourselves used the Foundation and the Information Modeler to prebuild very common PDM solution components. The Document Manager provides a basic document-management (or vaulting) capability, and it provides life-cycle management. A companion module to life-cycle management coming in Release 2 of Windchill at the end of 1998 is workflow management. This will allow users to automate the process of product development, to proactively take information and push it to a given user who completes a process with the information and then sends it on to another user. I would call life-cycle management passive process-management tracking, whereas workflow management is proactive process execution.

The Configuration Manager provides product structure management, view management, and change management. View management is an interesting subject from an enterprise standpoint because, depending upon who you are, you may have a different perspective on product structure. For example, an engineer might take a system view of the structure-what does the wiring system look like, what does the power transmission system look like, what does the heating, ventilation, and air-conditioning (HVAC) system look like? In the case of the HVAC system, the person assembling HVAC components doesn't assemble all the ducts with respect to each other. In fact, maybe what he does is assemble the ducts into the infrastructure and then assemble the sections together. So the assembly structure could look quite different from the design structure, and there should be some way of communicating that to the user.

ME: Many mechanical engineers with management responsibilities are assessing their current PDM solutions. What criteria should they use in deciding whether it will support enterprise PDM?

JH: From an engineering management perspective, I think there are a couple of key issues. First, is this solution making my engineers more productive and more effective within this local environment? Second, is this solution leveraging the investment we're making in my department to the benefit of the rest of the company? It's a two-part question, and sometimes the best solution locally may not be the best solution for the rest of the company, and vice versa. Because Pro/ENGINEER has a major presence in a lot of manufacturing companies, Parametric Technology Corp. is well positioned to propose solutions to the first question. Because Windchill is a Web-based technology that is easily exposed to the masses, PTC is also well positioned to propose solutions to the question of how to leverage the engineering investment across the enterprise.

ME: How would an engineering manager determine if engineering information was being leveraged across the enterprise?

JH: In some cases it may be completely obvious because there is no strategy for leverage across the enterprise. What user communities beyond engineering are validated to access this information? Are there any? Is there any activity from these user communities? These kinds of questions shift the focus from managing CAD data to improving the business through leverage. These questions tend to involve the information technology (IT) organization, because when a lot of departments feel they have a vested interest in the system, IT tends to be the organization that owns a cross-departmental enterprise system. As we move more from the Pro/ INTRALINK focus to the Windchill focus, we tend to talk more to IT people. Engineering managers need to be part of that conversation.