Disclosure: This post is sponsored by Datacor and reflects their views, opinions, and insights.
Process simulation has come a long way from pencil, paper, and rules of thumb, and AI is poised to change it again.
We spoke with Steve Brown, VP of R&D at Datacor, who has worked with CHEMCAD for decades. He shared where AI will actually make a difference in simulation, why engineers need to stay in the loop, how Datacor's engineering products are coming together, and the trends he expects to shape the industry in 2027.
You've worked with CHEMCAD for decades. What problem were you trying to solve then, and is it still at the heart of what CHEMCAD does today?
That question gets at the heart of what chemical engineers do in the process industries. We take low-value feedstocks, convert them into higher-value products, and make money in the process. Before computers, we relied on pencil, paper, and rules of thumb. As computing power increased, we could take fewer shortcuts and solve bigger problems.
But the shift has been about more than just faster, more accurate calculations. Today, engineers are solving for multiple objectives at once: maximizing profit, reducing energy use, reducing environmental impact, maximizing asset utilization, and more. Process simulators sit at the heart of that optimization.
AI is everywhere in engineering software right now. Where will it actually change process simulation, and where is the hype running ahead of reality?
Right now, the biggest impact is behind the scenes. Our product and development teams are using AI tools to work more productively, from understanding existing code to designing, coding, and testing new features. That means more functionality and reliability for our users.
On the user side, we're identifying friction points where AI can help. Building a flowsheet, for example, involves many repetitive setup steps that don't call for much engineering judgement, so we see real potential for AI to take on more of that work. Other wins include narrowing option sets like thermodynamic models, equipment configurations, and convergence strategies, along with context-aware help systems and powerful visualization tools for complex data.
Where I see hype is in training AI on datasets that represent real, physical systems without physics guardrails, or relying on self-learning systems to supply those guardrails. I'll admit I'm worried about that.
Simulation depends on rigorous thermodynamics and first principles. How do you make sure AI-assisted tools add to that rigor rather than wear it down?
This requires care, and I don't want to present myself as the expert. I've listened to many experienced engineers, managers, and executives in the industry who share this concern. In a recent talk, I said companies ought to challenge their vendors and demand four things: audit trails and transparency around AI reasoning, physics guardrails, clear data ownership policies, and integrated safety and reliability at every step.
Could AI eventually build or troubleshoot a flowsheet from a plain-language description? What role would the engineer play in that world?
Tools like that exist in various forms across the industry today, and they will continue to improve. But I think it's a mistake to take the engineer out of the equation. Engineers should check flowsheets as AI builds them, and any AI troubleshooting should result in a list of suggested actions for an engineer to review. We work in an industry that is dangerous by nature, and mistakes can cause significant damage to equipment and people. I want engineers vetting every step.
CHEMCAD now sits in your engineering portfolio alongside Fathom, Arrow, Impulse, xStream, GoldSim, and MySep. What does tighter integration look like for an engineer?
Having one company provide multiple products reduces some friction, but that isn't enough in today's engineering software space. The computing environment is finally ripe for a more modern approach. We'll use cloud tools, AI, and modern APIs to do three things. Users get easier access to the tools they already love and smoother workflows between them. Managers get visibility into their teams' work and productivity. Executives get a concrete basis for business decisions built on engineering results.
Datacor's broader portfolio also includes ERP, formulation, LIMS, and MES. Do you see simulation connecting to the business and plant-floor side?
In a word, yes. Customers often have three or more data sets about the same physical process: a simulation, sensor data from SCADA, DCS, or PLC systems, and lab data. We're working to let you analyze all of it at once, which can lead to better understanding of simulation gaps or plant performance issues.
Another path forward is proactive, advisory AI that draws on ERP, plant sensor, and simulation data. Executives could get a heads-up, based on their feedstock and fuel contracts and plant data, that their margins will come under pressure, along with simulated scenarios for addressing it.
Last year we talked about MATLAB integration. Which other external connections matter most going forward, and tell us about the partnership with Cleopatra.
We often learn from our users which tools they find most useful alongside CHEMCAD. Our goal is out-of-the-box compatibility with third-party software, and modern APIs go a long way toward that. Whether it's MATLAB and Simulink, Python, SCADA systems, or data historians, robust APIs will give users the flexibility to link CHEMCAD to a world of tools. Making that a reality is a focus of the CHEMCAD roadmap.
As for Cleopatra, customers told us they need tools that handle equipment and construction costs alongside detailed financial analysis for their projects. We're excited to partner with Cleopatra to bring CHEMCAD's simulation power together with their costing tools.
Looking ahead to 2027, what are the top industry trends you think will shape process simulation?
The workforce gap has been looming, and I hear about it from customers, at conferences, and in trade publications. We're losing experienced engineers, and engineering staffs are smaller than ever. Powerful engineering software won't solve that alone, so I look forward to working with our customers on features, possibly AI-driven, that help address this challenge.
I've seen many trends come and go, but a few have legs. Electrification affects refinery slates, which in turn affects the chemical industry. Water is another to watch, from access to clean water to use and reuse within facilities. And industrial biology, where yeast and bacteria make chemical precursors and products, is often less carbon-intensive and can run at lower temperatures and pressures than traditional processing.
What excites you most about CHEMCAD's next chapter, and what do you want customers to know about where it's headed?
I'm most excited about CHEMCAD's new role within the larger Datacor family. Some of the world’s most talented chemical engineers have used CHEMCAD to design, troubleshoot, optimize, and operate remarkable processes. When I think about how we can free engineers to focus on the actual engineering work by reducing overhead and assisting with decision-making, I get very excited about the new, better, cleaner, and more efficient processes our customers will create.
Datacor, which includes CHEMCAD, Flow Simulation products, GoldSim and MySep, is working toward simulation that is more accessible, better connected, and AI-assisted, with engineers firmly in control.

Steve Brown is the VP of R&D in Datacor’s Engineering Solutions Group. He helped define and grow the CHEMCAD business as a privately held entity—holding roles in technical support, sales, marketing, and management—and continues to lead the division after the 2021 acquisition by Datacor. Learn more.
Disclosure: This post is sponsored by Datacor and reflects their views, opinions, and insights.