Systems Engineering in Product Development

It’s not a new idea. It’s just another step toward a goal, which, says Chuck Grindstaff, president & CEO of Siemens PLM Software (plm.automation.siemens.com/en_us/), is to “define not only the traditional connections between all of the components in a product, but also its functional, safety, and sustainability characteristics in a way that lets us close the loop with the requirements important for that product.”

Siemens has a name for this engineering process: system-driven product design (SDPD). SDPD is akin to systems engineering, which the U.S. Department of Transportation defines as “an interdisciplinary approach and means to enable the realization of successful systems. It focuses on defining customer needs and required functionality early in the development cycle, documenting requirements, then proceeding with design synthesis and system validation while considering the complete problem.” 

SDPD, admits Grindstaff, is “really broad,” but it’s invaluable at sorting out the inherent complexity in today’s products, such as automobiles, by invoking a continuous process of decomposition, iteration, and closed-loop feedback through design, test, validation, manufacturing, and related processes.

Many challenges exist in implementing effective SDPD. In design, connecting the system of systems that make up an automobile. In production, connecting car design to the distributed and increasingly intelligent devices in manufacturing that make the car. In information technology, creating the infrastructure that supports distributed systems thinking; distributed design, simulation, analysis, and test; and all the distributed decision-making that goes with all that.

PRODUCT DEVELOPMENT EVOLVES

Vehicle development has changed over the years. It’s faster. It encompasses electromechanical components, software, and all the related connections, interactions, and behaviors of those elements—the convergence known as “mechatronics”—as well as magnetic, hydraulic, thermal, and other domains in physics. There’s also the realization that “requirements are part of product configurations, and should be embedded in the design process. They should be treated the same way as any other design deliverable,” says Stefan Jockusch, Vice President, Automotive Industry Strategy for Siemens PLM Software.

Innovative car designs, let along mechatronics itself, breeds complexity. Sure, many vehicle systems are natural and obvious: chassis, suspension, powertrain, body design, and so on. Breaking these systems down—a process called “decomposition”—into individual subsystems and components from an engineering perspective is also natural and obvious. Unfortunately, that breakdown does not account for all the multi-domain interactions across all the sub-systems—“or not as effectively as auto designers would like,” says Grindstaff.

PLM HELPS SDPD

Siemens NX covers three areas of product development, explains Jockusch. It’s geometry design. It’s the connection to simulation. It’s computer-aided manufacturing. (NX with Siemens Tecnomatix handle the digital-factory side of product design.) These software programs create massive amounts of data about geometry, simulations, analysis, and manufacturing execution. But, says Jockusch, “Engineers spend close to 60% of their time not doing engineering. 

They’re looking for information and aligning information, aligning requirements to attribute goals, and aligning attribute goals to technologies.”

This, too, is not new. Information management has always been a Holy Grail in product design and production. But therein lies the new idea: product lifecycle management (PLM) can be the foundation for all the software tools in design, analysis, simulation, visualization, control, and so on. That is, PLM enables SDPD. According to the product literature, Teamcenter, the PLM system from Siemens PLM Software, includes a consistent process and information infrastructure, cross-attribute simulation and test-based validation, configuration management, issue change and schedule management, and traceability throughout the product lifecycle—all based on open standards for hardware, software, and data integration. Grindstaff says all that much better: “Customers can take this ‘backbone infrastructure’ and link it to their analytical tools, whether the tools are from us, our partners, or our competitors.” PLM turns raw data into something people can quickly look at and understand in relation to the whole product lifecycle process. 

“The big deal is that [these integrated software tools] work at the scale of a car,” says Grindstaff. “The computers are fast enough, memory is cheap enough, the algorithms are robust enough, the simulators are accurate enough, the collaboration infrastructure is scalable enough. That wasn’t the case even five years ago. We can now capture the requirements, functions, and logic; manage them with configuration change control; tie them into the detailed designs; hook the simulations together through that data infrastructure; and work in more than just a `PowerPoint-y’ way. We now have real engineering.”

FROM PRODUCT DESIGN TO MANUFACTURING CONTROL

PLM’s ability to capture data and the rules related to those data applies to the design of manufacturing systems as well. Design and manufacturing engineering teams, says Grindstaff, “now have a common information fabric to communicate their decisions and results.”

The benefits are many. PLM-fueled SDPD, says Jockusch, can capture and represent the detailed know-how that experienced, veteran automotive engineers have, readying automakers for the upcoming generation of engineers. 

Product design and manufacturing engineers can try out all sorts of complex designs in a fully functional, virtual world that matches the physical world—before committing time and energy to a particular prototype. The virtual world can confirm that a car design will act as anticipated, and whether it meets the requirements for that new design. Design properties and rules can drive the work instructions and the factory devices that produce the cars. Last, decision makers are better able to make—decisions.

“People shouldn’t view this as a pipe dream,” concludes Grindstaff. “It’s something they can build right now.”