Testing many design ideas for a part, such as the suspension arm, means simulating many variations. Those with the most promise can be pushed along the development path.
Testing many design ideas for a part, such as the suspension arm, means simulating many variations. Those with the most promise can be pushed along the development path.

Thomas Curry
Executive Vice President
LMS International
Livonia, Mich.
www.lmsintl.com

Business has never been more competitive. There is global competition, heightened customer expectations, quality demands, product proliferation, and less time to earn a profit. These challenges require manufacturers make more and better products faster and at lower cost. All this calls for an innovative product-development process.

One route to innovation operates a design program that encourages fast, low cost, and accurate prototyping. Innovations don't just happen. They can come from a process that generates lots of ideas, the best of which must be screened and refined into commercially successful products. In the Harvard Business Review article "Enlightened Experimentation: The New Imperative for Innovation," Stefan Thomke describes how Toyota and others are able to identify and solve problems earlier by front-loading the product-development process.

At some point you need prototypes that can be shared with customers, tested, and evaluated. Evaluating ideas calls for a system to manage hundreds or even thousands of simulations at every level from components to subsystems to full product.

Four ideas

Creating an innovation system produces a few challenges. They involve developing ways to run simulations quickly, fostering a culture in which people aren't afraid to fail, front-loading designs with engineering effort early in the cycle, and blending new technologies with traditional methods.

Speed up simulation. It's essential to run simulations quickly so results can be used to improve the design. Simulation has often been out of step with design schedules. Results came too late to be useful. This created a lot of frustration, waste, and confusion as to what the simulation department should be doing.

To speed things up, models must have enough detail to meet required accuracy goals. Sufficient computing resources must be in place. And most importantly, the organization must be able to quickly get data into the right hands. This could mean redesigning the way things are done, such as major changes to processes, attitudes, and technology. It may also mean reorganizing the way groups work together.

Don't be afraid to let design ideas fail. Failures can have positive affects. When simulations are fast enough, even absurd ideas can reveal useful information. This is especially true early in the design process to eliminate potential losers and focus effort on potential winners.

Companies that stigmatize failed designs often have engineers building elaborate, expensive models to increase their confidence. With all that effort put in, they become defensive about designs and reluctant to explore alternatives. Worse, other people in the process may become reluctant to suggest changes, even if they see better possibilities.

Companies known for innovation, such as 3M, have created an environment where people aren't afraid to fail. Lockheed's "Skunkworks" had a similar reputation.

Front-load product development. Most product cost is designed in by decisions made early in development. Fixing problems later in the cycle (especially after releasing a design to production) can cost millions. Focusing engineering effort on a project as early as possible heads off such problems by getting designs right the first time.

In addition to saving time, early prototypes and simulations let companies do beta testing sooner to keep in touch with customers and react faster to shifting market requirements. Also, early simulations can significantly guide conceptual design. There is probably no greater potential benefit than defining a product's target performance to meet expectations of a particular market segment.

Combine traditional and new approaches. New methods can usually approach the same level of performance as traditional methods more quickly and at lower cost. For example, virtual prototyping digitally simulates many phenomena faster and at lower cost than traditional physical testing. Of course, some virtual predictions are just not as reliable as physical testing. Such testing may also be faster and easier to perform when existing components are available. But by combining new and old ways, engineering teams can avoid gaps new methods alone introduce and still benefit from speed and economy.

Leading-edge companies combine virtual prototypes with physical-test results into hybrid models. By combining traditional and new methods, we hope to eliminate the gap and actually increase the potential for innovation. In the case of automotive noise, for example, this means being able to rapidly explore alternative virtual subframes or engine mounts in the context of a real-world vehicle.

The payoff

The justification for implementing simulation technology is usually an ROI based on savings. In the automotive industry, for instance, virtual prototyping was initially integrated into product development as a way of improving the use of expensive test rigs. The greater benefit of the equipment is that they result in innovation, both in product designs and the design process. Enterprises are recognizing that the most sustainable and difficult-to-imitate competitive advantage comes from innovative products that define brand value and clearly differentiate the company.