Picture an automotive production floor. It's alive with activity. Materials move from cell to cell, advancing along a complex network of synchronized conveyors. Inside the cells, robots work at a frenetic pace, drilling, pressing, positioning, assembling, welding, and painting.
Lost amid the action, however, is the tremendous amount of work put in by production engineers to make everything happen on cue. For months prior to bringing up the line, they run back and forth between the drawing board and shop floor, designing cell layouts, programming robots, checking for collisions, and optimizing cycle times so production lines work smoothly and quickly.
Now imagine all these tasks automated by computers. With a few mouse clicks, production engineers access the latest product designs, call up virtual equipment from software libraries, and design and program the optimal cell without putting pencil to paper or stepping outside the office.
Need to modify to an engine component? In a computerized "digital" factory, it's no problem. Change the CAD data, and the manufacturing process updates almost automatically. Need to run a line faster? The machine models will help you optimize tool paths as well as motion profiles.
Until a few years ago, the notion of a digital factory was far from reality. Although some of the pieces were in place, such as CAD tools and computerized production machinery, no one had yet seriously addressed the need to automate the production engineering process.
Production engineering, the link between product design and manufacturing, is generally regarded as the main bottleneck in industry today. It's a slow process where there's little reuse of previous engineering. What's worse, production planning typically doesn't start until the very end of the design phase.
Recently, however, carmakers have begun to turn their attention to this neglected area, where many believe lie the biggest productivity and efficiency gains yet to be achieved in automation. Like they did in the 1970s with CAD/CAM, auto makers are beginning to invest heavily in computer-aided production engineering (CAPE) tools and practices. With this technology, they hope to cut down on prototypes, optimize production lines, and ensure relatively trouble-free start-ups.
Among the many carmakers on the road to the virtual factory are Chrysler, Ford, General Motors, Honda, Mercedes-Benz, Nissan, Toyota, and Volvo. And they're not bashful about it. In fact, Chrysler even promotes the concept of the virtual or "digital" factory in its current advertising campaign.
General Motors also sees a lot of promise in the new technology. "As computer-aided engineering spreads from GM's design process to its vast manufacturing operations in coming years, the months of development time will quickly fall from the 30s to the mid-20s and even lower," says Kenneth R. Baker, GM's vice president of R&D.
"To speed product development, GM is converting its vehicle design and testing from the realm of solid materials – where engineers actually have to build prototypes in order to test them – to a virtual world where engineering, design, and testing will all be done on computers," says Baker.
Making it work
Converting to the "digital world," as GM and others are trying to do, is not without challenges. Consider this: The typical automotive industrial process includes defining manufacturing features for an incredibly complex design; planning the sequence of operations; assigning resources; designing layouts, tooling, and tasks; validating and optimizing everything possible; and creating programs and manufacturing documentation. Even with CAPE tools, there's a lot of work to be done.
The place to start is with the design of individual manufacturing processes. Such processes may be robotic or manual, and include spot welding, fixturing and tooling, arc welding, painting, drilling, and riveting.
The first step in designing a digital factory is to create a graphical representation of the various workcells. A typical cell would incorporate several virtual machines and production equipment models.
Once the cells are organized into virtual manufacturing lines, they can then be manipulated to perform onscreen manufacturing activities. Within this virtual environment, production engineers can design, visualize, simulate, and optimize almost any manufacturing system.
One area where new technology is already having an impact is in digital prototyping. Using "dynamic digital mock-ups," engineers are catching product errors earlier than ever. They're also working out insertion and extraction paths for parts, and verifying service and maintenance procedures before the first real prototype is ever built.
CAPE technology is also being used to improve quality and ensure that parts are manufactured according to design intent. These advantages stem from the software's ability to see that tolerances are correctly specified during design and followed during the manufacturing and assembly processes.
Automotive engineers are also using the technology to program robots and other machines off-line. The software helps by generating numerical control solutions with collision- free toolpaths. What's more, using solid part models and information stored in memory, CAPE software can prescribe the best manufacturing process plan based on the cutting tools, machines, and methods at hand. If something changes, the plans automatically update.
In essence, CAPE software captures the factory in digital form, allowing it to "go back into inventory" for reuse on the next car model. The information regarding manufacturing resources is stored in a process database, or repository. This knowledge base typically contains the geometric representation and behavior of machines, tooling, processes, and even biomechanical models.
These and other benefits of CAPE and the digital factory are likely to become even more important in the future. As product-design life cycles continue to shrink and manufacturing operations become more costly and complex, remaining competitive in the automotive market will require manufacturing systems that outlive several product life cycles. Flexibility will be the key, and the virtual factory may unlock the door.
Putting the pieces in place
The foundation for the virtual factory was laid in the early 1980s with the development of CAD robotics. Using computeraided design tools, manufacturers began simulating the behavior of robots and so assessing the feasibility of various manufacturing tasks. This soon led to off-line programming, the use of computers to generate and verify software for numerically controlled (NC) production machines.
By the end of the decade, software suppliers were busy developing applications for individual processes, such as assembly, machining, quality control, arc and spot welding, and 3D laser and water-jet cutting.
The next big step, which occurred in the past few years, was to automate the production engineering process. The result of this undertaking is a new type of tool called computer-aided production engineering (CAPE) software. Properly installed, the software becomes part of a company's information technology infrastructure, communicating upstream with CAD tools and downstream with production equipment controllers.
With software extending across design, production, and manufacturing, the concept of a "digital factory" is now a reality. Today, engineers sitting at computer screens can create models and simulate the entire manufacturing environment. In this virtual world, they can explore and evaluate different industrial strategies in their quest to optimize quality, productivity, throughput, and return on equity.
Douglas Fish is general manager of North American operations for Tecnomatix Technologies Inc., Novi, Mich.