Paving the Way towards CAE Democratization
Consider for a moment all the technology people use on a daily basis without giving it much thought. Imagine that in order to extract any value from it, it is necessary to be an expert in all the physics behind it. This has been true for most products and solutions early in their development lifecycle before reaching the mass consumer. Take anything from photography to GPS—initially, new technology tends to be highly demanding for the user, allowing only a narrow circle of experts to benefit from it. Specialized expertise is often not the only obstacle—it generally also requires costly equipment and time-consuming maintenance, which makes it a questionable investment for professionals or hobbyists whose time, money, and training scope are limited.
This has been the case for Computer-Aided Engineering (CAE) solutions as well. And while most technologies eventually enter the so-called democratization stage—becoming affordable and user-friendly enough for an average user to comfortably operate it—this process has been exceptionally slow for CAE. Until recently, it seemed nothing could rock the stable CAE software market, dominated by a handful of major players offering traditional on-premises software solutions. This has been the industry standard for the past 50 years, however, the cost of licenses and hardware, as well as the knowledge and experience necessary to use these solutions, put them out of reach for most product designers in smaller companies.
How Can We Challenge the Status Quo and Enhance CAE Democratization?
The democratization of any technology is possible on two conditions: a drop in price and a lowered learning curve. CAE is no exception. Let’s consider the possible solutions in detail.
Lowering the Software and Hardware Costs
Pioneer companies like SimScale and Onshape have already made large strides in achieving this by harnessing the power of the cloud and introducing a flexible pay-by-use system. Most existing simulation tools are bound to a user’s desktop, and while for some simulations a few cores might suffice, more complex projects are impossible without access to high-performance computing (HPC). A cloud-based solution eliminates that problem allowing the user to perform the most sophisticated analyses using state-of-the-art simulation methods directly from the browser on a simple tablet, while having access to the required amount of computing hardware remotely.
The struggle with separate licenses for different types of analysis is a thing of the past as well. With monthly or yearly subscriptions, users can access all features of the platform, which in the case of SimScale includes simulation functionality ranging from computational fluid dynamics and thermal analysis to solid mechanics. The SimScale pricing model allows the user to start with simulation for free through the Community plan or the Professional account’s 2-week trial and scales the simulation capacity as it is needed, only paying for the cores that are used. This option opens up CAE to smaller companies which cannot afford to commit to long-term investments with large fixed expenses and high maintenance fees.
Aside from cutting the costs, the collaborative nature of a cloud-based platform can democratize the environment within companies as well. The fact that a simulation project is 100% accessible from a simple web browser and that anyone within the company can easily share their work with co-workers, partners, vendors, or customers can lead to the spread of knowledge and diffusion of decision-making. As a result, more engineers are let into the exclusive “inner sanctum” of CAE previously only accessible to a small number of experts.
Perhaps the only thing that is still holding some skeptics back from embracing the cloud is the security concerns; new product design often involves highly sensitive and confidential data after all. But while this psychological barrier experienced in some of the more conservative circles is understandable, experts say it is expected to melt away much like the skepticism around online banking did years ago. Data transferred to and from the SimScale platform is protected with industry standard encryption technology, using external data centers, as well as hard drives on the company’s servers, which are encrypted using the industry standard AES technology—which also protects the security of government documents. The communication between the user’s computer and the SimScale platform is done via SSL encrypted connections, a technology that is used in online banking and shopping. More about the security measures of the SimScale platform is available on this page.
Bridging the Knowledge Gap between CAE Experts and Product Designers
With that, the first condition of CAE democratization can be crossed off the list. The second one, however—minimizing the expertise expected of the user—, is not as simple. Product reliability is crucial to any design, which therefore must be tested within a large number of operating conditions. For each type of analysis and calculation, expert knowledge is needed to get reliable results. Smaller companies simply can’t afford to hire enough engineers capable of using the traditional CAE tools to perform all the simulations that are required to ensure that they are telling them what they really need to know.
Due to the extremely complex mixed-fidelity and multiphysics calculations involved—as well as the highly manual and non-intuitive user interfaces—CAE has been regarded for years as something akin to alchemy, the secrets of which would require years of intensive training to uncover. Product designers, however, want to focus on the product, and not on the tool. This is referred to as design-centric workflow rather than simulation-centric—the simulation should be just a means to an end and its primary purpose is to drive design-related decisions. Ultimately, many of the currently dominant CAE tools that are supposed to shorten the development cycle and make the process more efficient, end up slowing users down.
This barrier is far harder to overcome and unfortunately, there are no “quick fix” solutions here. It takes dedicated effort on behalf of CAE solution providers to educate their users and minimize the time they have to spend adjusting to the tool before they can get solid results. The measures that have proved to be most effective include:
Automation, Integration, and Templates – Performing a complex analysis from scratch takes time and sophisticated training that most non-experts do not possess. A tool becomes much easier to use when it is highly automated, templatized, and combines multiple functions. Integrating multiple state-of-the-art solvers (such as OpenFOAM, Code_Aster, YADE, and SU2 in case of SimScale) into the simulation platform provides a full set of simulation features in one single tool that can significantly reduce the manual workload. In addition to that, there are now more than 20,000 high-quality simulation projects in the SimScale Public Project Library, freely available for its community members to copy and use as templates.
Intuitive User Interface – While powerful CAE tools capable of performing complex simulations with intricate settings can only be simplified to a certain extent, a user-friendly, intuitive, interface can go a long way to make it easier for product designers to learn. However, simplifying the interface without sacrificing the tool’s functionality, or adding new sophisticated features without overcomplicating the interface is not easy, and SimScale is trying to maintain this delicate balance with its regular platform updates.
Learning materials – In the end, no matter how simple and automated the tool is, in order to get reliable results and not just a pretty picture, some background knowledge and training are necessary. Readily available learning materials, as well as interactive webinars and workshops, can significantly shorten the time required for product designers to get the hang of the tool without being slowed down by it.
It can at times be difficult for the tool developers to put themselves into non-CAE expert’s shoes, however, to reach a true CAE democratization, it is essential that experts focus on capturing their expertise in the form of templates and rules, and educate non-experts to interpret them correctly. One of SimScale’s own engineers described her personal experience learning to use SimScale for the first time and gathered a few tips that can be helpful for beginners—read her blog post to learn more: How to Learn SimScale in 30 Days.
CAE Democratization Going Forward
In conclusion, it must be acknowledged that CAE democratization is still an ongoing process that is far from completion. While the barriers associated with the costs of simulation software and HPC have been all eliminated by the introduction of cloud-based solutions, skepticism regarding security concerns persists. Since reducing CAE solutions to primitive apps is out of the question, the second obstacle—the expertise and knowledge gap— is likely to never be overcome completely and would take dedicated effort on the part of both—the tool providers and users. At the same time, there is no denying that a number of emerging companies have made significant progress in advancing CAE democratization. While there is still room for improvement, the success stories of hundreds of smaller companies and startups which would not have been possible without simulation prove that CAE democratization is a cause worthy of pursuing.