Congratulations! You've decided to join all the best-in-class organizations that utilize engineering simulation as an integral part of product development. So what happens next? There are lots of ways to get started, and the best method for implementation depends on many factors. Here are a few essential ideas that I've found to be helpful throughout the many years I've been working with companies on various aspects of CAE simulation.
A key factor to success is the CAE software (FEA, CFD, electromagnetics, etc.) which is chosen, but that's a topic unto itself. Let's begin with the assumption that the proper CAE software and support team have been selected, and the engineers have been trained. Where do you go from here?
Pick a project. The biggest obstacle in the introduction of CAE simulation is not technical; it's overcoming the human intellectual inertia, or our natural inclination to return to what we've done in the past. So, rather than simultaneously introducing simulation into the development process of every product in your portfolio and trying to overcome the resistance of a multitude of project managers and engineers at the same time, pick one product or initiative to start with.
The product you chose depends on how confident you are in success, your tolerance for stress and risk, and how badly you want to impress people at your organization. If you want to hit it out of the park, find the most high-profile project you can think of. All eyes will be on you, and if you're successful, you'll be a hero. On the other hand, if things don't go so well, it could get rough. Personally, I like to go for doubles -- mid-profile projects, with reasonable lead times, where you have a good handle on the existing product development process and timeline. Whichever project you choose, it's important that it be a good candidate for quantitatively demonstrating the benefits of simulation.
Identify bottlenecks and develop the simulation plan. Once the project is chosen, the next step is to examine the existing design process and identify the major bottlenecks. Select those points in the process where simulation can help grease the skids. Next, form a detailed plan for how simulation will be used and make the appropriate modifications to the design process. This plan must include the individual responsibilities of cross-functional teams, the types of simulation to be conducted, the level of simulation required, the expected timeline, and defined criteria for an acceptable solution.
It is critical to ensure that there's a way to measure the impact of this plan on time and cost. This is how you're going to convince others that simulation is beneficial. Management cares about numbers, so you need to provide data and not starry-eyed descriptions revolving around the "cool factor" of seeing the videos of the simulation results. Identification of the bottlenecks and developing a plan isn't always intuitively obvious, so you may want to get assistance from an outside expert. This is the most important part of the implementation and one you can't afford to get wrong. Remember, this is a numbers game, and you're looking to demonstrate that the numbers are better with simulation than without.
Communicate the plan. If you're the only one who knows about the simulation plan, this isn't going to end well. It's important to pull all the people together who have a vested interest in the project's success and explain to them what you're trying to accomplish, what their responsibilities are, what will be measured, the required timeline, and how they will benefit from all of this.
The final and arguably most important point is to monitor, measure, and adjust. In a perfect world, you just need to set the plan in motion, go away for a six-month sabbatical, and come back to view the wonderful benefits of simulation. Now, let's get back to reality. The chance that everything will go smoothly on its own the first time out can make winning the lottery look like a sure thing. It's important to carefully monitor the process to ensure that the plan is both being adhered to as well as working as intended. Try to measure what can be measured at every stage. Remember, simulation plans are dynamic, so make modifications as appropriate and in real-time. Don't wait for a post-mortem; it might be yours.
At the end of all of this work, you will have:
- A better product, released with time to spare
- A quantifiable measure of how simulation improved the product development process
- Lessons learned, and a better understanding of how to apply simulation to the next project
- An established simulation-centric communication process among the cross-functional team members
- A healthy raise (or at least an "atta boy")
— Nicholas M. Veikos, Eng.Sc.D., is president of CAE Associates Inc.