RSS Feeds
Management vs. Luck
Data management gains importance as design complexity soars
Terry Costlow, Contributing Editor -- Design News, October 8, 2006
When Donald Gradin is designing escalator and elevator replacements being installed in 58 countries at sites as large as O’Hare airport, he’s always concerned that all the engineers are working with the latest version of the plan. If someone misses a change made in R&D centers in India, Germany, Italy or headquarters in Finland, errors could be costly.
The mechanical engineer in charge of escalator modernization for Kone Inc. is pretty blunt about how he’s avoided problems that can occur when updates don’t get passed to everyone. “Luck,” he says.
Gradin, based in Coal Valley, IL, laughs when he says it, but updating constantly changing files is a serious problem. Most engineering managers worry about the impact a small oversight can make during a complex project. Development cycles are shrinking while complexity is soaring, factors that team up with outsourcing to create an environment where keeping everyone on the same page is getting increasingly difficult.
When many engineers work on a project, the chance that a single disk drive failure could derail the production schedule is always present. “Now, all our engineers have models on their own computers. If they’re good, they back them up,” says Brad Bransen, manager of furniture engineering at Williams-Sonoma, a Sonoma, CA retailer, that designs more than 2,000 products per year. In systems with thousands of elements, a tiny glitch can have big consequences.
“It takes us about 14 months to do a production line, and we generate around 14,000 files. It just takes one person to accidentally move one file to create havoc,” says Dave Gadberry, engineering manager at Can Lines Engineering, a production line integrator in Downey, CA.
With more engineering staffs managing hundreds of projects and thousands of files, relying on luck is no longer a viable solution. Software suppliers too, realize that engineering managers probably don’t want to add “data manager” to their title, but they don’t have many alternatives.
“Data management is a non-optional aspect of running an engineering department. It’s a necessary evil,” says Bob Merlo, marketing vice president at Autodesk Inc. of San Rafael, CA.
He notes that the move to 3D files plays a big role in the need for data management. “You have seven times the number of files when you go from 2D to 3D,” he says.
Rippling Back
Though a number of companies address various aspects of file management, the field is still wide open. The many facets of managing engineering files that continuously change make it difficult for vendors to come up with solutions. Systems must be transparent so engineers in remote sites don’t have to issue storage commands.
Management tools must also account for manufacturing engineers who make decisions that can ripple back to design. “You definitely want communications to be bidirectional, so any changes made in manufacturing are seen in engineering,” says Lance Murphy, product marketing manger at the Enovia Division of Dassault.
Data management systems must also integrate with the many tools engineers use.
Can Lines used an independent data management system to manage files created by design teams in Downey and Appleton, WI. But pressure to shorten development times has prompted them to switch to a storage system that is tightly integrated with the development software the teams use. “It took nine to 10 months for our data management company to catch up with new tools. We couldn’t use new features during that time,” Gadberry says.
Integration is a key aspect in many design tools. Dassault has a broad line, so its tools have techniques to make sure everyone connected to a program knows about changes that impact them. “When an author in Catia makes a change that impacts manufacturing, we create a notification so manufacturing engineers can act on it in a timely fashion,” Murphy says.
Design for Manufacturability
Much of the focus is on making sure all design engineers know what their counterparts are doing. But for the many companies that have design and manufacturing teams working in parallel, software tools must also account for communications that reverse the normal flow.
In fields like aerospace, where manufacturing engineers often develop equipment while designs are still in flux, decisions made in manufacturing can ripple back to the design team. For example, manufacturing engineers may use some drill holes to hold a wing segment in place as it moves through the production line.
“If the design engineer doesn’t know that some holes are now tooling holes, he may change something and make the tooling obsolete. When manufacturing engineers augment files with data about their tooling holes, that will flow back to the design engineer, who will know that changing those tools will be extremely expensive,” says John O’Connor, market strategy director at Vistagy Inc. of Waltham, MA.
Automated change management is particularly important in regulated environments like aerospace and health care.
“Our design cycles are very fast. Seventy percent of our models change every year because we integrate the best PCs, monitors and vital sign management equipment. Many of these systems need agency approval, so everything has to be saved,” says Darin Janoschka, director of product development at Flo Healthcare of Norcross, GA.
Janoschka adds that using a single repository also adds some standardization that facilitates reuse, a critical factor for most of today’s companies. It also makes it easier to compare designs. “Having casters show up in the same place on every bill of material is a huge benefit,” he says.
Keep it Simple
Though engineers generally agree that data management is becoming a necessity, most don’t have a ton of time to implement schemes, nor do they have a huge budget. Those factors have prevented broad usage of product lifecycle management software, which was supposed to perform data management across the entire enterprise. “PLM deserves its bad rap for low return on investment,” Merlo says.
Techniques that focus on engineering are much easier to implement than corporate-wide schemes that must span many departments. Janoschka started a two-user test at the start of the year and felt confident enough to expand it to the full team by the end of summer.
Merlo notes that standard data formats are an important factor for success when different engineering disciplines are involved. “If you don’t have a common data format, it’s difficult to share data,” he says.
Others note that systems must be able to accept changes that are made well after designs are complete. Manufacturing teams often require changes that the design team must implement.
“Ideally, you want to provide a design that can be altered in response to customer requirements and when there are manufacturing changes for lean manufacturing or Six Sigma,” Murphy says.
The focus on data management is getting good reviews from engineering managers who don’t want to worry about sending e-mails and using other techniques, including luck, to make sure everyone is updated. “I’m really glad we’re getting some tools for this. It will make life a lot easier,” Gradin says.
Get more information on Autodesk’s data management software.
Gallery »
Management vs. Luck
Data management gains importance as design complexity soars
