When you hear the term “Product Development System” (PDS), you might think it's just a cool PTC marketing slogan. When you take the time to learn more about it, you might realize that it's a fundamental requirement for operating today's global businesses.
At ITT Corporation, we decided to embrace PTC's Product Development System concept in 1998 and its current form in 2004, adopting the parts that are most applicable to our business needs. This was the beginning of an interesting and often exciting journey that many of today's organizations are contemplating or have already embarked upon. While not without problems and setbacks, creating a PDS has had huge payoffs for our customers, supplies, and employees alike.
ITT is a global, multi-industry company with primary businesses in the areas of defense electronics and services, fluid technology, and motion and flow control. In 2005, ITT employed 44,000 people in hundreds of offices around the world, most of them outside North America. The enterprise comprises three management companies. Within each of these companies are “Value Centers” or groups that develop, create, and manufacture products for a specific market segment.
A Little Bit of History
Our PDS initiative built on the progress we had already achieved in creating “single-source data”—a common, integrated set of product information required to support our corporate strategy to “Design anywhere. Build anywhere. Sell, service and integrate everywhere.”
ITT first articulated its strategic vision in the early 1990s, well before the infrastructure was in place to support such an ambitious information strategy. To maximize the tools available at the time, we in large part made our businesses and business processes ERP-centric and used email as both a workflow and content management tool.
When 2000 rolled around, we attempted to leverage various connectors for email, ERP and the internet. Unfortunately, the information that was such a cornerstone for our efforts generally remained hidden. The lack of true workflow and knowledge management tools and platforms made it difficult to demonstrate the value of single-source data and get the necessary buy-in from businesses and users.
In 2001, new "spaces" like the Collaborative Commerce Model (CPC), driven by the dot-com boom, came into play. Although short-lived, this proved to be an important milestone because industry as a whole was beginning to accept the product data strategy that ITT was championing. To us, CPC was based on core processes (product development, demand creation, order acquisition, etc.), which in turn were built on basic infrastructure processes (business-to-business, business-to-consumer, network management, etc.) and managed with governing processes (strategic planning, capital allocation, budgeting, etc.) The glue holding it all together was connectivity and free information flow across the enterprise. The key concept here is that technology is the enabler and process redesign the foundation.
Figure 1. End Point Vision Diagram
Single-Source Data
In business today, information is a valuable asset. Information lying dormant or unreachable is as costly as excess inventory sitting on a shelf in a forgotten warehouse. An information infrastructure that enables the enterprise to "Design anywhere. Build anywhere. Sell, service and integrate everywhere.” must therefore start with digital data repositories for customer, vendor, supplier, OEM and internal product information.
To accomplish this, we needed to standardize on a common set of master tools that define our “product” interface for all transactions while leveraging a shared knowledge base without translation. By minimizing translations and hence errors (defects), we improve quality and customer satisfaction, as well as reduce overall costs and the need to rework. Furthermore, using an integrated set of tools from design to aftermarket also improves data integrity and enables design re-use, common sourcing, and common part numbers, components and materials.
Common tools facilitate the standardization of Value Centers with common processes. At the same time, the Value Centers can leverage each other's data while automating processes where possible and where it makes the most sense. Again, this prevents defects and shortens cycle time as we integrate various disparate systems. We have also realized real cost savings by negotiating with a single vendor as a corporation vs. individual businesses working with select suppliers. Moreover, common sense and experience taught us that integration of data is more efficient when standardized with common formats. There are also the added benefits that knowledge can easily be shared amoung users of common tools.
The most important payoff, however, is that this approach helps us extract additional value out of our internal manufacturing capabilities, while at the same time leveraging our significant capital investments. In addition, single-source data and a common data platform allow us to
meet continuous process improvement goals (VBSS – Value Based Six Sigma)
meet our organic growth goals (VBPD – Value Based Product Development)
improve asset utilization across the enterprise, and
implement e-commerce initiatives.
The Tools
1. Pro/ENGINEER
Standardizing on Pro/ENGINEER laid the foundation for our product data strategy. As most of you know, Pro/E is a parametric, feature-based, fully associative database that is platform-independent. This powerful software provides a variety of capabilities for engineering our products by creating detailed solid and sheet metal components, building assemblies, producing fully documented production drawings and photorealistic images, and manufacturing models—all in virtual 3D space.
The key reason why we chose Pro/E over competitors is that it is based on a single, integrated data model. We are not a software company and programming interfaces between various software companies' solutions is something we do only without choice. Reducing the integration between various functions was a priority, and PTC had already done this for us in the area of mechanical design software.
2. Pro/INTRALINK
One of the first steps was to implement Pro/INTRALINK at each major business location. Back in 1999, this was the best technological solution available for managing complex product data like Pro/ENGINEER files. Based on our strategy and vision, we made the decision not to use Pro/INTRALINK to manage all product data because of our expectations for Windchill (with which we had no experience at the time) and concerns it would make upgrades and eventual migration to Windchill much more problematic.
The former Engineered Products Solutions Group (EPSG), which was already developing products collaboratively using manual methods, was chosen as the first multi-site Pro/INTRALINK implementation. These sites were manufacturing the same product to meet both US and EU standards. To expand colloboration outside a few multi-site servers, users manually pointed a Pro/INTRALINK client at a different remote server. While employees could thus collaborate beyond normal boundaries, this approach was still not optimal.
It didn't take long before the limitations of this form of enterprise collaboration became obvious. As the EPSG businesses expanded and products were shifted to different design and manufacturing sites, the stress on users increased dramatically. They had to be constantly aware of where to locate the data, how to manage it while working on it, and where to put it back when finished.
What we faced were lots of Pro/INTRALINK systems doing a great job at managing Pro/E data but unable to handle global needs. ITT had implemented PTC's ProjectLink in 2001 to help support our collaboration and program/project management.
Figure 3. Pro/INTRALINK Topology (circa 02-04)
3. Windchill
In early 2001, ITT was the first company to complete a production implementation of ProjectLink. In 2002, after a series of successful exploratory pilots, we followed this up with our first production implementation of a Windchill Foundation-based system for our global engineering change system, along with other supporting processes. As PTC listened to feedback from us and other customers, it became clear that we needed a set of preconfigured, integrated applications to reduce the cost and time of implementation. We needed out-of-the-box functionality that addressed real business needs. The result was the expansion of the “Link” applications, including PDMLink.
Fast forward to 2004, our worldwide team selected PDMLink and ProjectLink as the base components of what we call the Global Vault. We envisioned this as a platform that would meet our needs in three main areas—Vaulting (Search, View and Markup), Enterprise Change Processes (Graphical Workflow and Lifecycle Management), and PDM (Pro/ENGINEER Data Management).
The Global Vault would eventually replace ProjectLink as a standalone application, Windchill Foundation, all of our Pro/INTRALINK 3.x servers, and legacy systems such as Access databases, shared network drives, etc.
Global Vault as Platform
The Global Vault has four major components: create (Pro/ENGINEER), collaborate (ProjectLink), control (PDMLink), and communicate (Arbortext).
1. Create. ITT has millions of pieces of Pro/ENGINEER data dating back to the late 1990s. Although it is easy to focus on the cost of a particular CAD system, the real value is in the data. As such, data created in Pro/ENGINEER Release 5 in 1990 is still accessible in Wildfire 2.0 in 2006. Not every CAD system can make that claim.
Standardization on a single CAD system from concept through design and manufacturing has clear advantages. The fewer systems used to create data, the better because it significantly reduces translations, which in turn reduces errors and enables more streamlined processes. (This is as true for a word processing system as for a CAD system.)
2. Collaborate . We now use an integrated ProjectLink as part of the PDMLink environment, right out of the box. We also leave a great amount of flexibility to the project creators/managers and end users for how it is used. The rule is that ProjectLink should be used for all collaboration and projects related to product data. Other collaboration tools, such as our ITT portal, are for collaboration and projects related to non-product data.
3. Control. Rather than some uber-ruler that controls everything, the Global Vault integrates with systems that are channels to the intranet, internet and extranet. The Global Vault is focused on controlling, managing, sharing—through the use of document management, workflow, lifecycle management—all product data. Internally, there are still business and other systems geared to specific purposes that may either obtain information from or provide information to the Global Vault.
4. Communicate . At ITT, documentation is sometimes as critical as the product or service itself. Efficiently and accurately providing this information, in different formats and channels, is currently a costly and time-consuming process. In some cases, communicating too much is not always a good thing. For example, a user guide that is delivered with an engineered product that details options that are not included in the “As Shipped” configuration can confuse the customer and result in a less than optimal experience.
Focus on Need
We have boiled the requirements for the Global Vault down to three specific needs:
Share product information globally
Share product information securely
Create a single source of product information
These criteria are the first gate a concept must pass before it becomes a project. For instance, if a team comes forward wanting to access information in the Global Vault for the purpose of dumping it into a website with no access controls, the request would be summarily rejected. It is important to note that these are needs. The functions around consuming, sharing and accessing the information are different for specific businesses, suppliers or customers.
Making the Vault Truly Global
“Communicate” is the latest capability added to our product development system. Introduced by our largest Value Center, ITT Flygt, the Arbortext global enterprise publishing project may be the most important initiative at ITT today. Dynamic publishing can be likened to mass production of documents from a common database of discrete components.
Figure 4. PTC PDS schematic for products, the ones shown are in production or are in the advanced implementation phases.
Why enterprise publishing? Well, think about it like an engineer. What if every Pro/ENGINEER assembly was made up of features in one big part file (like when you import an IGES or STEP file into Pro/E from another CAD system). How difficult would it be to find, let alone re-use, a set of geometric features? Could you operate efficiently with Pro/E if that was a process you had to follow? The question, then, is why accept this as business as usual in our everyday documentation processes?
Figures 5a, b, c. ProE Model 1, 2 and 3
In the traditional publishing process, each author creates and manages both the format and the content. The content is transferred manually into different publications and documents using email, cut and paste, etc. These documents are in turn converted to different media, such as high-resolution PDFs for print and HTML for the web, to communicate through different channels. Under this system, technical writers spend hours updating hundreds of individual documents every time engineers make minor changes to a component used in different products. Authors also spend a lot of time searching for things they had written or translated, and half of the time their searches are unsuccessful.
Figure 6. Independent Publication Process
With Arbortext, in contrast, discrete “blocks” of standard information are mapped to different document types and formats, including websites, CD-ROMs, PDFs and wireless devices. The assembled information is automatically formatted to suit the needs of the channel. Now, making a change is much simpler because it affects only one single block of information.
Figure 7. Single Publication Cycle
Automatic updating of the parts list is one of the first dynamic publishing applications to go live, providing an immediate savings of some 2,000 man-hours a year. The parts list is now configured and updated by retrieving the part numbers and descriptions from the mainframe, the exploded view of the assembly from the CAD system, and a short explanation typed into a web application and stored for reuse.
Figure 8. Arbortext Linking Documents
The ITT Flygt project has expanded into a Fluid Technology Management Company project with all four Value Centers (Flygt, RCW, IBG and AWT) working collaboratively. Arbortext will ultimately integrate a translation memory database with the Windchill PDMLink component of our Global Vault, which will constantly “learn” from previous versions and allow a very selective update when the basic text is changed. This will save an enormous amount of time and money because it means that information will be translated just once and again only when changes are introduced.
But this is only the tip of the “benefits iceberg” from using the Arbortext software. Once our dynamic publishing system is fully implemented, more and more of our documents will be generated and updated automatically. At that point, our technical documentation will be accurate and up to date, more accessible and user-friendly, and presented with a consistent ITT look and feel.
What's in Store?
Is the system perfect? No. Are there issues? Yes. Implementing a Product Development System is tough both procedurally and technologically, not to mention culturally. The good news is we learn from every problem and we continue to make it better. Even if the challenges sometimes seem insurmountable, the value that PDS brings to a global enterprise is many orders of magnitude greater.
You have heard the saying, “Life is not a destination, it's a journey.” Well, there's still a long road ahead with many more obstacles to overcome before we have a fully functioning product development system. But when we look at how far we've come, we are more committed than ever to face whatever challenges lie ahead.
Figure 9. Global Vault Data Process Design
Ron Watson is global product data manager at ITT Corporation and a member of the PTC/USER Board of Directors. This article is based on his keynote presentation at the PTC/USER World Event 2006.
