Putting People and Processes before Technology (PLM Strategies Column)31 Jul, 2007 By: Kenneth Wong
Consultants discuss creativity, control, and collaboration.
At the recent CATIA Operator Exchange (COE) conference that took place at Las Vegas' Rio Hotel (April 29–May 2), the deputy chief information officer of an automotive supplier revealed that innovation might be too rich for his blood.
"If you let engineers innovate, they will innovate — by God, they will," he exclaimed, rolling his eyes. His weary tone suggested he'd had more than his fair share of innovation. The other IT executives seated next to him chuckled and nodded to show they shared his pain. One added helpfully, "Innovation is fine, but somebody has to pay for that."
In this article
The lunchtime chatter and the clinking of silverware didn't obscure this IT chief's underlying message: He was looking to rein in what he considered to be rogue engineers running amuck with wild ideas. And he and the other IT leaders could very well do it with PLM tools. After all, they were at a PLM conference. So what does that say about PLM as an innovation driver?
Betting on Creativity
"Some PLM components do have to do with control," said Ed Miller, president of CIMdata, a PLM consulting firm. "They have to do with data vaulting, permission control, process control, and so on. With tools like that, there's always the opportunity to overcontrol. When you do that, you stifle creativity."
The correct approach, Miller recommended, is to use PLM "to automate the mundane stuff engineers are forced to work on. When they're not using up their time with those, then they can work on what they're paid to do — engineering." For overworked executives, the temptation to use PLM to control, or even limit, the flow of imaginative product ideas will be difficult to resist. It'll take forward-thinking executives with a long-term vision — the ones who are willing to take a chance on new ideas — to promote creativity with PLM tools.
Miller said that a certain vehicle released approximately three years ago stands out in his mind as the classic example of an engineering gaffe. Much to the relief of the auto manufacturer, he said he couldn't remember the exact brand and model. But he had this to say about its design:
"It looks terrific, but in order to do one of the most routine maintenance tasks, you have to remove one of the engine brackets, because you can't get the servicing tool into that gap," he said. Miller said he often recounts this cautionary tale to emphasize the importance of collaboration. "If someone, perhaps a service technician, had looked at the design from this perspective, the flaw would have been caught; it would have been a relatively trivial design change."
The latest generation of PLM products offers collaboration functions aimed at preventing such blunders. Recognizing a stellar market, graphics giant Adobe dove into the engineering collaboration space. With the May release of Acrobat 3D v8, "CAD, CAM, CAE, and technical publishing professionals in the aerospace, automotive, consumer electronics, heavy machinery, life sciences, and AEC industries can convert virtually any 3D CAD file — including large assemblies," Adobe announced. "Three-dimensional CAD data can also be easily combined with other critical project information, such as product specifications, spreadsheets, and bills of materials, into a more secure PDF document containing product manufacturing information [PMI]."
David Prawel, a 3D software expert and founder of the consulting firm Longview Advisors, said, "BMW is one of the companies leveraging the power of 3D in downstream applications to enhance the quality and effectiveness of repairs. I was one of the people involved in a project where it developed a 3D maintenance and repair application that assisted mechanics in performing some common repairs using 3D manuals and animated repair instructions." According to Prawel, BMW found that these repairs were easier for mechanics to complete, more understandable, of higher quality, and very well accepted, even by highly experienced mechanics.
Illuminating the Blind Spots
"In the recent years," said Miller, "we've seen a huge focus on executive dashboards. They give you a snapshot of your project. So, if you're a project manager, you use certain matrices to determine what's going on in your project, where you might have problems, where you can apply your time most effectively, and so on."
Centric Software, which describes its offerings as product intelligence applications, delivers this function in its Centric Decision Center, a graphical interface for displaying aggregated project statuses derived from live data that resides in ERP (enterprise-resource planning), PLM, PDM (product data management), SCM (supply-chain management), and other legacy systems. According to the company's online literature, it lets users
- 1. visualize multilevel KPI (key performance indicator) metrics and current status for projects, products, portfolios, and programs
- 2. identify critical issues as they arise and take immediate action
- 3. promote strong projects and suspend failing ones
- 4. track the effectiveness of decisions and programs
At PTC's World Media event in June, Brian Shepherd, the company's divisional vice-president of product management, introduced Windchill 9.0, which features business reporting functions from the third-party software vendor Cognos.
"[Our customers] embed a lot of intellectual properties and knowledge in Windchill [PTC's data-management platform]," said Shepherd. "This new module allows them to get reports, get insight into their operations — where are the problems, where are the opportunities for improvement? These are not static reports. They're interactive, so they can do what-if scenarios."
At COE, Dassault Systèmes' president and CEO Bernard Charlès practically leapt with excitement as he introduced what he called "the first of a broad branch of Dassault Systèmes' online application family." 3DLive, a 3D PLM environment for navigating PLM repositories (figure 1), promises (among other things) "a single immersive interface that connects the people, processes, and products needed to accelerate decision making and drive innovation."
Figure 1. Dassault's 3D Live, introduced at COE, gives project managers the ability to quickly visualize the product components alongside live manufacturing data. This screen shows the breakdown of a vehicle structure.
An Explosive Lesson
"The technology is not a magic bullet," Miller cautioned. "It's there to facilitate the processes, so it really falls back on your processes." It also relies on the decision makers' expertise and wisdom to make the right choices.
In 1977, Mother Jones magazine reported on the Ford Pinto, which contained a fatal design flaw ("Pinto Madness," September/October 1997). An investigation by the magazine revealed that "Ford engineers discovered in preproduction crash tests that rear-end collisions would rupture the Pinto's fuel system extremely easily . . . Because assembly-line machinery was already tooled when engineers found this defect, top Ford officials decided to manufacture the car anyway — exploding gas tank and all — even though Ford owned the patent on a much safer gas tank." The bottom line: No executive dashboard can prevent poor judgment.
In his keynote speech at COE, Prawel reiterated a point he often makes: "To implement effective PLM, companies need to think about people, supporting processes, supported by technology. Too much software is acquired before core business processes and workflows are thoroughly understood, synchronized, and standardized."
Six years ago, engineering powerhouse Delphi figured out it had invented something worth safeguarding. It wasn't a product; it was an idea, a series of standardized processes that governed the 3D CAD modeling workflow (figure 2). Evidently, the discovery was important enough to warrant intellectual property protection. Delphi uses the terms horizontal modeling (HM) and digital process design (DPD) to describe the steps involved. Today, in Delphi's official papers, HM and DPD are followed by trademark symbols; they're now patented Delphi processes available for licensing for a fee.
Figure 2. Delphi's horizontal modeling (HM) principle replaces the traditional sequential engineering workflow and the associated parent–child relationships with a more flexible simultaneous production process, made possible by standardized modeling methods.
Prawel was one of the people involved in the HM-DPD initiative. Establishing a uniform set of modeling practices is critical groundwork, he said, "because you can't automate something in software unless you have standardized the processes you're trying to automate."
Kevin Marseilles, a senior process designer at Delphi, found himself spending countless hours dismantling the CAD models submitted by designers just so he could produce the process drawings that show step-by-step how the product was supposed to be machined and assembled on the shop floor. In the worst cases, he'd end up tossing out the designers' 3D models and recreating them from scratch.
"The problem was," he said, "the [feature-based 3D solid] models supplied by the product groups were difficult to break into individual processes." That was because there was no standard CAD modeling methods. The designers produced the features needed in their own ways, relying on their schooling, training, and preferences. Now, with standardized feature modeling practices in place (that's HM), the process drawings can automatically be generated (that's DPD), often with minimum intervention from Marseilles.
Codified Behavior for the Masses
In a company whitepaper, Delphi printed a breakdown of the benefits derived from the initiative. After implementation, Marseilles and other process designers saw a 50% reduction in the time spent to create — or recreate — the CAD models and 90% reduction in the time spent editing them ("Horizontal Modeling & Digital Process Design," Delphi).
In a Delphi Web seminar titled "Delphi Design Methodologies: Improving Global Design Collaboration for Manufacturing Excellence," Marseilles demonstrated HM and DPD in Unigraphics NX and Team-center software environments. Afterwards, fielding questions from attendees, he said, "[The Delphi process] works with Unigraphics, CATIA, SolidWorks, Solid Edge, and probably every other CAD system that we know of, but we haven't tested them all yet . . . [It] works very well with both [manufacturing products internally and externally]. It keeps us concurrent. The data is clean, so we don't have the usual problems we used to see in the older days."
According to Prawel, "In the end, PLM comes down to coding a set of processes in a piece of software. How can you code your processes if you don't have standardized processes?"
An ideal PLM suite, in Prawel's view, "is a PDM system that's very open." He predicted sometime in the near future, perhaps in the next two or three years, "somebody is going to implement SOA [service-oriented architecture] correctly. That will function as the backbone to share disparate data, a well-defined interface that links all the islands of automation. That's an effective infrastructure to enable PLM, but you won't have PLM until you have standardized the processes that you want to implement."
Prawel keeps a blog called 3D Ubiquity (www.3dubiquity.com) where he comments on industry trends. Ed Miller responds to readers' questions in Innovate Forum's Ask the Expert (www.innovateforum.com) series.
Kenneth Wong is a former editor of Cadence magazine. As a freelance writer, he explores innovative usage of technology and its implications. E-mail him at Kenneth.Wong at cadalyst.com.