Executives’ Reflections, Hopes, and Cautions for 2025 and Beyond

To open the year 2025, we approached industry veterans and leaders to cast their eyes wide and far on the design, simulation, and additive manufacturing industries, and share their thoughts. Here are some (with minor edits for length):

Worries and Concerns for the Future of Design and Simulation

I worry about the misperception that simulation software is too complex or inaccessible, which prevents many organizations from adopting digital engineering. Modern simulation tools enable faster, cost-effective transitions from concept to production by replacing traditional, time-intensive processes with advanced digital methodologies like multiphysics, automation, and AI ...

To unlock its full potential, simulation must be integrated earlier in the design process, during system architecture development, while fostering collaboration between software suppliers, hardware developers, and end-users. Workforce training is equally critical to ensure engineers can effectively use these tools.

—Dr. Larry Williams, Distinguished Engineer, Ansys

With the diversity of simulation tools and influx of new technologies, it can be a challenge to keep channel partners’ teams of experts trained and knowledgeable across the simulation spectrum.  Additionally, the proliferation of many smaller players claiming to offer the “next best thing” in simulation has the potential to create confusion in the marketplace.

—Jason Pfeiffer, VP, Rand Simulation

With disconnected, disparate products, organizations can only achieve incremental productivity gains. To see breakthrough productivity gains, data must flow seamlessly and be connected end-to-end. The productivity increases will be a welcome accelerant in and of themselves, but they’re also the fuel for building more AI-powered automation tools.

—Jeff Kinder, EVP, Product Development and Manufacturing Solutions, Autodesk

Many organizations lack a centralized platform to collect and manage data that is vital for training AI models, which may introduce potential risks such as inaccuracies in automated designs, lack of transparency in decision-making, and potential security vulnerabilities that require careful management. Without a clear picture of all the data at your disposal, AI models also won’t function to their highest capability. This includes know-how in addition to hard knowledge. Key learnings are critical to input into AI models to complement facts to ensure that all users do not make the same mistakes twice.   

—Manish Kumar, CEO, SOLIDWORKS

Worries and Concerns for Additive Manufacturing (AM)

What concerns me about the future of additive manufacturing is the risk of limiting its full potential due to a lack of openness. Many AM companies still operate in a closed ecosystem—restricting access to materials, machine parameters and interoperability between systems. This protectionism keeps prices high and prevents manufacturers from optimizing their production processes to meet their unique needs. To truly empower customers, the industry must break down these silos, offering more flexibility, broader material choices and seamless integration across different technologies. At Materialise, we believe openness is key to driving innovation, which is why we’ve taken steps like opening up our Magics software to enable custom workflows tailored to specific manufacturing challenges.

—Brigitte de Vet, CEO, Materialise

Hypes to Fact-Check and Verify Before Embracing

We can all agree that AI is revolutionizing engineering design by automating tasks like optimization, meshing, and generative design, while leveraging historical data to improve efficiency and decision-making ... However, challenges remain, particularly around data labeling. For example, autonomous vehicle AI requires labeled data to identify objects like cars and pedestrians. Synthetic data generated in simulations can help, as it is pre-labeled, but it is most reliable when grounded in real physics.

Similarly, simulation data within engineering organizations holds enormous potential for training AI, but unstructured or unlabeled datasets often require significant preparation to make them usable ... For example, retraining on AI-generated data has led to bizarre outcomes, such as nonsensical chatbot responses. To avoid these pitfalls, organizations must use accurate, physics-based simulation data as a ground truth and establish guardrails to prevent contamination of source data. 

—Dr. Larry Williams, Distinguished Engineer, Ansys

While tools like ChatGPT showcase AI’s potential, simulation AI remains complex and will take quite awhile to reach its full potential ... Some of the product development community, while excited by simulation AI’s potential, are concerned about its current level of maturity. This is reflected in the caution exercised by some large companies, which are explicitly prohibiting their simulation suppliers from incorporating AI into the simulations conducted on their products.

—Jason Pfeiffer, VP, Rand Simulation

Distributed manufacturing has long been hailed as a major benefit of additive manufacturing. However, we see too much hype when 3D printing companies try to apply this model to many different industries at once. Doing so often doesn’t work because each has its own nuances and very different quality, post-processing and certification requirements. We see the most success when a 3D manufacturer focuses on one industry and/or builds competence in a certain application. The aerospace industry is a great example of how this can be successful, with 3D printed replacement parts filling supply chain gaps by printing parts on demand in a distributed manner to return aircraft to service more quickly than with traditional manufacturing.

—Brigitte de Vet, CEO, Materialise

Under-Appreciated or Overlooked

Engineering simulation is that it is fundamentally a tool—an amplifier of an engineer’s skill and creativity—rather than a substitute for human ingenuity. Just as a hammer doesn’t make a carpenter, engineering simulation doesn’t create engineering brilliance; it enhances the expertise of individuals who have invested years mastering their craft. This foundational human effort is sometimes underappreciated.

Another underappreciated aspect is the untapped potential of engineering simulation to solve challenges or develop entirely new ideas without ever building a prototype ... Imagine a scenario where an existing product must be redesigned to be 50% lighter. Instead of spinning up an entire engineering team, automated procedures could leverage simulation to explore solutions using newer, lighter materials or by choosing generative organic shapes for lightweighting. In this instance, the engineering simulation can amplify the engineering effort of a much smaller team for the redesign, so long as the original product was built in a digital thread that can be leveraged. 

—Dr. Larry Williams, Distinguished Engineer, Ansys

While companies often invest heavily in simulation solvers and pre- and post-processors, they overlook the importance of HPC. Maximizing hardware core count, whether on-prem or cloud, dramatically reduces simulation solve times. This, in turn, reclaims enormous amounts of engineering time to enhance ideation or focus on other responsibilities beyond analysis. While HPC comes with a cost, when compared to the recouped engineering time and subsequent value from that outcome, the ROI is tremendous; especially for organizations using simulation on a frequent basis.

—Jason Pfeiffer, VP, Rand Simulation

While we are experts in 3D printing, our customers are true specialists in their fields—whether it’s aerospace, medical devices, eyewear or consumer products. Each industry has different requirements, and addressing them effectively requires more than just powerful AM tools based on our expertise. Instead, we need to develop solutions that fit seamlessly into their workflows, empowering them to leverage their own expertise.

—Brigitte de Vet, CEO, Materialise

Predictions 

I believe we’ll see a fork in the AI road. The fascinating and novel capabilities we’ve become accustomed to thinking about as defining AI, such as natural language prompts yielding fantastic images, essays, and code, will continue to advance. At the same time, very practical, somewhat mundane AI capabilities will emerge. And these advancements will save us immense amounts of time. Soon, busy-work tasks that used to take hours, or even days, will be completed with a simple click of a button. And we’ll get that time back to do the creative work that humans excel at, while the computer focuses on computing.

—Jeff Kinder, EVP, Product Development and Manufacturing Solutions, Autodesk

AI will be the determining factor that ultimately streamlines and optimizes design capabilities, but 2025 can be considered a bridging year to ensure that AI models have all the data in place to ensure organizations are maximizing their potential.  

—Manish Kumar, CEO, SOLIDWORK

More Ansys Coverage

More Autodesk Coverage