Every company wants to be innovative so they can deliver breakthrough products, Jim Cashman told attendees at the 2004 ANSYS International Conference May 24. Experimenting as a way to drive product innovation was one of the themes of the Management Track at the conference. But too many companies fear failure. Cashman says that simulation is one of the tools that can help them overcome that fear.

What are the major trends you are seeing in engineering these days? There are many, including globalization, the drive for more productivity, mass customization, and the move toward "green" products. Reduction in cycle time is a problem for all engineers. But one big characteristic of today's economy is that manufacturers are expanding their focus from manufacturing to product innovation.

What's the role of simulation in innovation? The role is definitely not to make pretty pictures on computer screens. Simulation provides a way of looking at alternatives quickly and investigating more alternatives than before in the same amount of time. Simulation is a form of experimentation, which leads to higher reliability in the new products you develop. Plus, it helps you get to market faster, a key aspect of innovation. In our opening management track session at the ANSYS conference, Harvard Business School Professor Stefan Thomke advised companies to fail early and fail often. I agree- and they should do it quickly, early in the process.

There has been a lot said about the importance of managing uncertainty. Does simulation help there? Yes, because it lets you consider a lot of alternatives early. But still, managers have to let their engineers know that it's okay to fail. Don't fear failure because failure can lead to innovation.

Will simulation ever replace CAD as the "paper napkin" for conceiving new products? Probably not. But it certainly helps in the concept phase. Engineers don't always need all the design details early in the planning phase. What they need to look at are design characteristics. Simulation lets the functional parameters drive the geometric parameters. For example, if you're designing an engine and you want a certain amount of power from that engine, looking at the flow characteristics before the packaging can be good. Simulation lets you do that.

Which is worse, no analysis of designs or bad analysis? No analysis is not an option today. You have to define "bad" analysis. Even 20 percent correlation is not predictive. It's indicative. It allows you to rank the alternatives. Without analysis, you're betting that you didn't make a mistake. Either you do physical tests that send costs through the roof, or you take a big risk. Simulation is the most cost-effective risk-management tool.

What is the ANSYS Workbench? It's an architecture that supports all our products, and a crucial part of our own internal efficiencies. We have been focusing on ways to make our products customizable, because no single product can be productive "as is" for everyone. Allowing customers to customize our products for their applications has been a key driver for us. Workbench ties our products together and helps that customization because it gives the customer just one tool that he or she has to maintain.