Case Study: The data builds the products
- 13 March, 2003 11:17
ITT Fluid Technology has been working on a project to support the global sharing of design, engineering, manufacturing and other product-related information among its various operations. The initiative, part of a wider effort to implement a "design anywhere, build anywhere" capability, should also let the manufacturer sell, service and integrate its parts everywhere, says global products and data manager Ron Watson.
The New York-based subsidiary of engineering conglomerate ITT Industries is one of many manufacturers turning to product life-cycle management (PLM) tools to improve manufacturing efficiency, product quality and time to market.
The systems let different groups exchange and collaborate on product-related images and data in real time, both internally and with outside suppliers and customers. To do this, the systems must gather and integrate data from sources ranging from design and engineering systems to manufacturing and field operations.
A wide variety of collaborative products fall under the PLM umbrella, including product data management technologies that integrate data from bill of material, computer-aided design (CAD), change management and enterprise resource planning (ERP) systems; configuration management systems that validate product configurations and link product specifications with engineering modules; and maintenance, repair and overhaul modules that link field services systems with change management and engineering.
While the benefits of linking such systems can be enormous, organizations that have launched PLM projects caution others to prepare for process and procedural changes, cultural issues and implementation challenges.
Because PLM systems are cross-functional in scope, installing one is often more difficult than companies anticipate, says Erica Rugullies, an analyst at Giga Information Group. The integration work involved in tying together disparate systems is substantial, she says.
"Companies can go about this in a few ways," Rugullies says. "They can look upon this as a huge project that requires extensive integration, software dollars and a number of players in different roles, or they can go after a specific business problem, such as finding a new way to effectively share digital data." But in either case, users should fully understand the issues involved in planning and deploying a PLM system, users and analysts say.
First and foremost, PLM products are still maturing, so the risk is that the software won't always live up to expectations. "PLM software can be buggy, it could require extensive programming to deliver the needed functionality, and it may not enable compliance with government requirements," Rugullies says. She recommends conducting a proof of concept before purchasing a PLM tool and budgeting time for troubleshooting and implementing fixes.
In addition, a full-fledged PLM project can be nearly as complex as rolling out an ERP system, so knowing upfront what value you want to get out of it is vital, says Marc Hendrickson, a design documentation manager at Team Industries, a manufacturer of powertrain and chassis components. Only then is it possible to know what information in a product's life cycle needs to be captured, how to manage it and what kind of workflow and decision-making processes need to be implemented. Team Industries uses Parametric Technology's (PTC) Pro/Engineer CAD and Pro/Intralink data management software to tie data from its parts-development process to shop-floor data. In doing so, Hendrickson says he hopes to improve quality, parts reuse and time to market.
Setting the right scope is important, says Jim Heppleman, chief technology officer at PTC, a vendor of PLM software. For instance, if the business goal is to reduce time to market, the project emphasis should be on enabling capabilities such as rapid digital prototyping and product bill-of-material management, Heppleman says. Similarly, if the goal is to increase product standardization, the focus might be on improving configuration management or on design reuse, he adds.
It also helps to have standard tools and interfaces when creating and sharing a single view of enterprise product data, says ITT's Watson. ITT is linking and integrating product design and engineering data across several manufacturing operations. To facilitate the process, ITT standardized on PTC's Pro/Engineer software for creating product design data and on Pro/Intralink to share data between sites.
Using a common software platform reduces the complex integration and data translation issues that are otherwise involved, he says. Using standardized tools can also result in better design integrity and reuse and lead to better sourcing and material management processes, Watson adds.
Taking Small Bites
When planning a deployment, don't bite off more than you can chew, warns Hendrickson. "In a small organization where resources are limited, it might make the best sense to take a phased approach," he says.
His company adopted a three-phase approach to implementing PLM. In the first phase, he "nailed down" the company's parts-development process in order to determine what type of data he needed to capture and when. (Team Industries has identified 50 separate data points as a result of this exercise.)
Team Industries is now in the second phase: tying the data to information generated on the manufacturing floor. In the third phase, it will integrate that data into the company's manufacturing resource planning system.
Customizing packaged software to fit specific business needs can also be a problem. New Balance Athletic Shoe in Boston is using PLM products from MatrixOne, to improve the effectiveness of its product development and procurement process and to make collaboration with international partners easier.
But New Balance has had to customize the software "a little more than we wanted to," says Bill Armelin, the shoe company's business systems project leader. For example, some procurement forms that the PLM system generated were different from what internal teams were used to, even though users input the same data. Similarly, what internal design teams refer to as a "specification sheet" appears as a bill of materials function in the PLM suite.
Be prepared to tweak your processes to avoid software customization, says Hendrickson. Team Industries changed its parts-numbering process and revisioning scheme to better use the capabilities of the PLM suite it's rolling out.
Getting users to adopt PLM systems can also be a challenge. "One of the things we noticed is that if you try to do too much right out of the box with a system like this, it is too much of a culture shock," says Tim Bahr, manager of technical development at New Balance.
Getting nonengineering people to use PLM systems can be especially challenging, so it's important to spend time explaining and demonstrating the value a PLM system can bring to the different business units, adds Watson.
His group conducted road shows at several ITT manufacturing facilities to demonstrate the value of the PLM system. But even with those efforts, Watson says that "only a very small percentage of people . . . really believe they need this stuff."
Equally important is the need to get user feedback and include that in the requirements-gathering process, says Rugullies. Putting PLM training groups in the middle of the units that are most reluctant to adopt PLM may also pay off, she suggests.
Ultimately, getting PLM right is as much about setting the right expectations as it is about the implementation, says Frank Azzolino, president of vendor Eigner. "This is an industry that has historically overpromised but underdelivered," he says.
Scrutinize Before You Buy
Given the complexity of the typical product life-cycle management (PLM) project, selecting the right product is key, say users and analysts.
Tim Bahr, manager of technical development at New Balance Athletic Shoe, says that when his company started investigating PLM, it had no idea where to begin. So Bahr and his associates visited eight companies that were implementing PLM technology and spent hours speaking with their peers about implementation issues and product options.
After narrowing the field to two, Bahr's team attended each vendor's introductory training session to get a feel for the technology before making a final selection.
This kind of extensive assessment is especially important given that a slew of vendors are rushing products to market. The more well-known players include PTC, Electronic Data Systems, IBM, MatrixOne, Agile Software and Eigner. ERP vendors such as SAP AG are also extending their technologies with PLM features.
There is no "right" PLM product that fits the bill for all users, says Ron Watson, global products and data manager at ITT Fluid Technology. Rather, product choice should be dictated by your current environment and needs. "It's easy to make a wrong choice," he says.
In ITT's case, Watson ultimately decided to go with Parametric Technology (PTC) because his company was already using PTC's Pro/Engineer software for its design work. That choice avoided interoperability problems, he says.
In the end, a product that is configurable to your processes is ideal, adds Erica Rugullies, an analyst at Giga Information Group. Otherwise, users could end up spending a lot of time customizing their software to make it fit in their environments, she cautions.