ASME.MVC.Models.DynamicPage.ContentDetailViewModel ContentDetailViewModel
Manufacturing Blog: Generative Design and Engineering Jobs

Manufacturing Blog: Generative Design and Engineering Jobs

CAD generative design boosts creativity, not productivity.
Let’s face it: good design is difficult. Really difficult. And with demands for more complex products, shorter schedules, and lower costs, it is only going to get harder.
These demands leave engineers with little or no opportunity for innovation. Fortunately, though, new and novel engineering tools could help you address these challenges. In this blog post, we’ll explore one of them: Generative design, artificial intelligence-based engineering software that has already begun to accelerate and disrupt the traditional design processes.
What is generative design?

Since engineering software is overrun with confusing acronyms and terminology, let’s get things clear right from the beginning.
Generative design is a set of algorithms that enable computer-aided design (CAD) or computer-aided engineering (CAE) software to autonomously generate the geometry of a design. As in, it creates or refines 3D geometry for you. On its own.
How does it work?

Generative design starts by defining the design goals, requirements, and constraints and inputting them into the application. These objectives and limits might relate to weight, cost, structural flexibility, or stress. They may also include manufacturability, such as defining whether you plan to mill or 3D print a part.
Once defined, the software generates a diverse range of design alternatives and iteratively “evolves” those that best meet your requirements. With each iteration, the algorithms test and learn what works.
Generative design algorithms are not standardized. They vary with each CAD or CAE application, and a single tool might use multiple methods.
For example, topology optimization is an established generative design algorithm that analyzes an existing piece of user-defined geometry and removes unnecessary material from the part. This makes it a subtractive generative design method.

Further Reading: How Industry 4.0 Impacts Engineering Design
Biomimicry, on the other hand, mimics behaviors seen in nature, such as the growth of tree branches and roots or the growth of bacterial colonies to optimize a specific design requirement like porosity or low-mass strength. This is an additive generative design method because material is added to the design.
Applications with generative design capabilities may produce a single design or many. Topology optimization, for example, generates one design as it removes material from the original. Biomimicry, however, provides variability against the same inputs and produces a range of possible geometries.
For the most part, designers and engineers use cloud-based computer platforms for generative design. The cloud gives your organization scalability and access to unlimited computational power at a low cost. Others prefer to use high-performance local networks or desktop computers.
How will generative design change modern design work? Does it boost productivity?

Foremost, think of generative design as a way to boost innovation. By autonomously providing a range of design options for review, it can help uncover novel design solutions to meet your product requirements. Early on, you can use it to compare and contrast design concepts. Late in the cycle, you
can tune geometry parameters like round sizes and rib thicknesses.
Unfortunately, there is a tradeoff: To achieve this, you need to devise the functional definition of your product. This is commonplace for electrical and electronic engineers, who use those definitions to autonomously route traces in a circuit board. This is not a commonplace methodology for mechanical engineers, and can prove to be a stumbling block.
Still, many mechanical engineers have been creating informal functional definitions in their heads for years. All they need to do is document those definitions as inputs. It is an essential undertaking, much like creating a logic diagram for a circuit board. Without this initial step, you cannot run traces on the board. Likewise, you cannot use generative design to create a mechanical fit-for-purpose product without defining its functions.

Further Reading: Software Bots: Will RPA Take Your Job or Help You?
This may sound like the generative design process is adding more steps to your design process. True, it takes time to input the functional definitions. But you will spend a lot less time manually modeling your designs because generative design will effectively do the heavy lifting for you. So, the net sum of work is the same as before.
You didn’t misread that. In my experience, generative design will not boost productivity.
So, why bother? I would argue that generative design helps us realize significant opportunities in innovation. This is because the generative design process may introduce a solution that’s outside the bounds of our imaginations. Its comprehensive abilities let engineers iterate many design possibilities to find the best solution without manually creating each permutation themselves. So, expedited innovation, rather than higher productivity, is where the real opportunities exist.
What is the biggest hurdle with generative design?

If you want to introduce generative design at your organization, the biggest challenge will come from your people. This is because, when people are short on time, they inevitably revert to known processes and practices.
Generative design is not only unfamiliar, but creating functional definitions requires more up-front work. It is only when the algorithm has been implemented and produced a design solution that you start to see results.
Also, companies need to invest in new technologies and staff training, which can be a roadblock for many businesses. It’s not as simple as installing the latest release of a CAD or CAE solution and getting everyone to use the new functionality. Organizations and their design teams need to invest time and money—two resources often in short supply—to reap the full rewards of generative design.
Those savvy organizations that do will see handsome returns on their investment, allowing them to expedite innovation across the development lifecycles of their products. But that works only if everyone at the organization is onboard.
Is generative design a threat to engineers’ jobs?

Everyone has seen articles warning that AI will take your job. Does that argument apply here?
Frankly, I just don’t see it.
If generative design offered clear-cut productivity gains, it might have been true. After all, making engineers twice as productive means management can halve the number of engineers. But, honestly, engineers that employ generative design have a whole new class of work to do. Without a dramatic boost in productivity, I don’t see it threatening engineering jobs.

Further Reading: The Future of Design

This does not mean it will never be a threat. In the future, someone might develop an AI-enabled voice assistant on the front end and an augmented reality display on the back end of generative design. Then, well, you’d have Jarvis from Ironman. But let’s be honest: Tony Stark isn’t walking through that door anytime soon. At least, in my opinion.
So, what’s the takeaway? Generative design can be a powerful tool in the modern engineer’s toolkit. But don’t expect it to work like magic. You’ll have to learn functional definition and change your design processes. Yet, given the increasingly complex challenges today’s design teams face, generative design could be the helping hand your organization needs.
Chad Jackson leads Lifecycle Insights and conducts research on technology-led initiatives for engineering executives.

Opinions expressed are the author’s and do not necessarily reflect the views of ASME.

You are now leaving