This Autumn, we are excited to release a suite of new features focused on automation, solver robustness, and expanded physics capabilities. From the introduction of our new Engineering AI Agent to major upgrades in our FEA and CFD solvers, this update is designed to shorten your simulation lead time and accelerate innovation.

Here is an overview of the key updates from the Autumn 2025 release.

On-Demand Webinar

If the highlights caught your interest, there are many more to see. Watch the on-demand Simulation Expert Series webinar from SimScale on how real-time simulation with AI is driving faster design cycles and superior products by clicking the link below.

Watch On Demand

SimScale AI: Agentic Automation

We are taking a significant leap forward in how users interact with simulation through our combined AI technologies: Engineering AI and Physics AI.

• Workbench Agent (Beta): The first step in our Engineering AI roadmap, the Workbench Agent acts as an intelligent assistant within the platform. It can help validate setups, answer simulation queries using documentation, and even automate workflows by reusing setups from previous projects.
• Physics AI Integration: Engineering AI can now leverage Physics AI to perform rapid design optimizations, allowing you to predict results instantly before validating the final design with a traditional solver.

To experience the power of Engineering AI and Physics AI together, click through the demo below. If you would like to take part in our beta program, contact your support team.

Ray – AI Chat Support:

Available to all users, Ray is our 24/7 AI support assistant. Ray can visually diagnose error messages from screenshots and provide instant troubleshooting steps, ensuring you are never stuck waiting for a resolution.

Ray is designed to be a SimScale expert. It has memorized our entire documentation and is trained on simulation best practices, as well as the most common issues and workflow questions you are likely to encounter.

Ray helps to get the information you need as quickly as possible, however your dedicated human engineer is still here for complex, deep-dive project support.

Computational Fluid Dynamics (CFD)

This quarter brings significant quality-of-life improvements and feature parity to our CFD solvers, particularly for rotating machinery and thermal management.

• Variable Time Steps (Multi-purpose): You can now define variable time steps for transient simulations. This allows users to start with coarser steps to establish flow stability and switch to finer steps for high-accuracy resolution, significantly optimizing runtime without compromising quality.
• Run Continuation for CHT (IBM): Users running Immersed Boundary Method (IBM) Conjugate Heat Transfer simulations can now continue transient runs or steady-state simulations from where they left off. This saves valuable computing resources by eliminating the need to restart long simulations from zero.
• Periodic Boundary Conditions: Efficiently simulate complex, repetitive structures (like heat exchangers) by modeling only a single unit cell or subset, reducing model size and computation time.

Finite Element Analysis (FEA)

Following the integration of the Hexagon Marc solver earlier this year, we have released a massive set of new capabilities to handle complex nonlinear structural problems.

• Load Steps: You can now define multiple load steps within a single simulation run. This is essential for manufacturing processes like pipe bending or rubber seal compression, where constraints and loads change sequentially during the event.
• Automatic Contact Detection and Glued Contact Tolerance: For large assemblies with 10+ parts, SimScale now automatically detects and glues contact pairs. This automation drastically reduces setup time, allowing you to focus only on the specific contacts that require manual definition.
• Contact Forces and Pressure: Get deeper insights into nonlinear simulations involving contact by reporting on forces, stresses, contact gap/state and contact pressure.
• Advanced Material Models:
  • Marlow Hyperelastic Model: Drive hyperelastic material behavior directly from test data without complex parameter fitting.
  • Viscoplastic Materials: Accurately model creep, relaxation, and damage in polymers under cyclic loading.
• Remote Displacement & Force: Apply distributed forces or rigid body motions (including rotations) via a central pilot point using new RBE2/RBE3 connectors.
• Symmetry Constraints: We have introduced a plane-based symmetry condition, allowing you to model only a half or quarter of your geometry to save computational effort while maintaining full-model accuracy.

Electromagnetics

Our electromagnetic analysis capabilities continue to grow, with a focus on high-frequency efficiency and power electronics.

• Litz Wire Modeling: You can now model complex bundles of insulated wire strands as a single equivalent geometry. This preserves key electrical and thermal characteristics while avoiding the computational heavy lifting of modeling every individual strand.
• Time-Periodic Acceleration: For transient magnetic simulations driven by periodic excitations (e.g., PWM or sinusoidal waveforms), this feature accelerates the solver to reach steady-state results in a fraction of the time.

General Platform Enhancements

We have also introduced several non-physics updates to improve the overall user experience and workflow efficiency.

• CAD Section View: Inspect the internal details of complex 3D CAD models by slicing them along a cutting plane directly in the pre-processor.
• Cancel CAD Operations: You now have the ability to stop long-running CAD operations midway, giving you better control over your preparation workflow.
• Automatic Extrusion Meshing: The Standard Mesher now automatically detects the optimal direction for extrusion (longest or shortest dimension), ensuring higher quality meshes for elongated bodies and thin plates.
• Pedestrian Wind Comfort (PWC) Cutting Planes: Post-processing for PWC analysis now supports cutting planes, allowing you to visualize comfort plots hidden by complex geometry like canopies or overhangs.

Watch Now

Don’t miss out on the full experience and deeper insights into how SimScale’s latest features can transform your engineering workflow. Watch the complete webinar on-demand to see these tools in action and understand how they can be applied to your specific challenges. Click here to access the webinar recording and start accelerating your design process today!

The post Webinar Highlights – SimScale Autumn 2025 Product Updates appeared first on SimScale.

But are we experiencing the same in mechanical engineering?

Our State of Engineering AI survey suggests that only 3% of hardware engineering companies are seeing comparable ‘significant’ gains from adopting AI in their workflows. Why such a small number? One of the most commonly cited reasons was the inflexibility of legacy software.

Luckily, there is a whole raft of new AI tools for mechanical engineers emerging, covering the whole product lifecycle. Let’s dive in.

1. Ideation: Beyond the Blank Page

The conceptual design phase sets the foundation for an entire project. Traditionally, this relies on an engineer’s experience and intuition to sketch out a few potential solutions. AI, specifically through generative design, challenges this paradigm.

Generative Design

Generative design tools use algorithms to explore thousands of potential design solutions based on a set of constraints you define. You input the non-negotiable parameters—functional requirements, material properties, manufacturing methods, and performance criteria—and the AI generates a massive number of high-performing options. Your role shifts from being a generator of a few ideas to a curator of many.

Autodesk Fusion 360

Autodesk’s platform uses cloud-based machine learning to automatically generate and rank design solutions. As an example, check out this case study with General Motors, which used generative design to redesign a seat bracket. The AI-generated result was a single, organic-shaped part that was 40% lighter and 20% stronger than the original component, which had previously consisted of eight separate parts.

nTop 

This tool excels in creating highly complex, performance-critical components. For instance, Cobra Aero used nTop to redesign a drone engine cylinder. Instead of traditional cooling fins, the software generated an intricate internal lattice structure. This AI-driven design significantly reduced weight while improving thermal performance—a result that would be nearly impossible to arrive at through traditional design methods.

2. Drafting, Design and Review: The Intelligent Drawing Board

Once a concept is chosen, it must be translated into a detailed CAD model. AI is now being embedded directly into CAD software to serve as an intelligent co-pilot, automating repetitive tasks and streamlining the design process.

AI-Assisted CAD

This category includes AI features that automate common tasks like dimensioning, applying constraints, design reviews and compliance checks. These tools learn from your design habits and company standards to make intelligent suggestions, reducing errors and saving significant time.

Onshape AI Advisor

This tool acts as an intelligent assistant directly within the cloud-native CAD environment. An engineer can ask natural-language questions (e.g., “How do I create a variable-pitch helix?”). The AI provides step-by-step recommendations, troubleshooting help, and best practices drawn from Onshape’s training materials, effectively accelerating the learning curve and day-to-day workflow. This cloud-native approach is also critical for AI-driven simulation, as it allows parametric CAD models to connect directly to analysis tools like SimScale, enabling the rapid, automated iteration that AI requires.

Shapr3D

Shapr3D bypasses the hours typically spent in dedicated rendering software with its embedded AI Visualization. By simply positioning a model and typing a text prompt (e.g., “modern kitchen counter”), engineers can generate context-aware renders in seconds. The tool uses your specific CAD geometry as a guide, allowing for rapid concept validation and “mood boarding.” Features like Variation Intensity let you control the AI’s creative freedom, enabling you to present polished, stakeholder-ready visuals instantly without leaving the design environment.

Intelligent Design Review (ReviewOps)

CoLab Software

While tools like Onshape and Shapr3D streamline CAD creation, CoLab addresses the bottleneck of reviewing it. Traditionally, critical feedback gets lost in “data graveyards” of static screenshots, email threads, and PowerPoint decks. CoLab’s AutoReview utilizes AI to automate the administrative heavy lifting of this process, scanning models against your specific design standards before a human ever looks at them.

This acts as an intelligent “pre-flight check,” automatically flagging routine errors like missing hole callouts or GD&T non-compliance. By filtering out this administrative noise, the AI allows senior engineers to focus their expertise on complex problem-solving rather than checklist verification. This effectively closes the loop between design intent and manufacturing reality, ensuring faster, higher-quality feedback cycles.

3. Simulation and Optimization: No More Waiting

For many engineering teams, this is the most painful part of the product development lifecycle. Traditional simulation (CAE) is a well-known bottleneck, but it’s actually two problems. The first is simulation cycle time: the hours or days you wait for a complex computation to run. The second, and often bigger, problem is the simulation lead time: the days or weeks spent by specialists manually setting up physics, preparing geometry, and re-meshing every new design variant. This ‘test-and-wait’ workflow means engineers can only analyze a handful of designs. Engineering AI is the solution to this entire bottleneck, tackling both problems at once to achieve one goal: no more waiting.

Engineering AI (Solving the Lead Time Bottleneck)

This category of AI acts as an intelligent co-pilot to automate the complex, multi-step setup process. Unlike a traditional macro or script which follows a rigid list of commands, Engineering AI uses Large Language Models (LLMs) to reason through the physics of your model.

SimScale Engineering AI

SimScale has introduced an agentic AI assistant that resides directly in the simulation platform. It transforms how engineers interact with simulation through three core capabilities:

1. Democratization for Novices: Traditionally, simulation required years of specialized training. Engineering AI lowers this barrier by guiding novice users step-by-step. It can diagnose missing inputs, suggest appropriate settings based on the geometry, and flag potential errors before a simulation is run. This turns the platform into a mentor, helping junior engineers get to valid results faster.
2. Promoting Best Practices: For larger organizations, consistency is key. Engineering AI can be configured to enforce company-specific “Gold Standard” settings. The agent ensures that every simulation—whether run by an expert in Germany or a novice in the US—adheres to the same quality standards and methodologies, reducing the risk of human error.
3. Reasoning and Adapting: Unlike rigid scripts, the agent uses reasoning to navigate deviations. If a geometry changes slightly or a parameter is missing, the agent evaluates the context to adapt its approach rather than failing.

Physics AI (Solving the Computation Bottleneck)

This is the surrogate model. It’s a lightweight, data-driven approximation of a high-fidelity simulation. An AI model learns the complex, non-linear relationships between inputs (geometry, boundary conditions) and outputs (performance, stress, temperature). Once trained, this ‘Physics AI’ model provides near-instant predictions, solving the computation time bottleneck.

SimScale’s Integrated AI Platform

SimScale has integrated this technology directly into its workflow. This approach was highlighted in a real-world case study by Convion, a part of HD Hyundai, who needed to optimize a complex hydrogen ejector pump.

The Challenge: Armin Narimanzadeh, Senior Thermofluids Expert at Convion, faced a multi-objective optimization problem. The design was complex, and using traditional CFD-driven optimization to find the best-performing design would have taken months.

The AI Solution: Using SimScale, Armin’s team first generated a training dataset by running a number of simulations. This data was used to train a reusable Physics AI model within the SimScale platform. The results were transformative. The team now has an AI model that can generate a new, optimized design in under an hour.

This “months to hours” transformation is a perfect example of AI’s power. It was made possible by a fully cloud-native toolchain: a parametric model built in Onshape was connected via an API to SimScale, allowing the AI to automatically test hundreds of variants to find the optimal design. The Engineering AI component automates this workflow, while the Physics AI component provides the instant predictions.

Learn more about the full process by watching the AI Engineering Bootcamp webinar (Session 1) on-demand.

4. Deployment: From Smart Manufacturing to Live Maintenance

AI’s role doesn’t stop when the design is finalized. It extends into the manufacturing process and the operational life of the product (deployment). Increasingly, smart manufacturing systems 

Smart Manufacturing

AI is streamlining the path from a 3D digital model to a physical part. CAM software is using AI to automate the complex and error-prone task of CNC programming, which is a major bottleneck due to a shortage of skilled machinists.

Siemens NX CAM

NX CAM uses AI for “Feature-Based Machining”. It automatically analyzes a 3D model and recognizes geometric features like holes, pockets, and slots. It then suggests the most suitable machining operations and sequences, even learning from a programmer’s past choices to improve future suggestions.

Real-time Digital Twins

Once a product is deployed, AI can help you operate it more effectively by building a real-time digital twin. A Physics AI model fed with boundary conditions from a live system can provide crucial insight into function and health, including ‘virtual sensing’ of metrics which might be impossible to directly measure. Machine learning can also be used to predict when it will fail before it happens, based on historical operational data.

Physical AI

Connect live sensor data streams from your operational assets or system-level models directly to Physics AI surrogates via SimScale’s open API. This creates a closed loop where real-world conditions continuously inform and update your digital twin for maximum accuracy.

By combining live data and hardware-in-the-loop (HiL) setups with AI models, you move from reactive maintenance to predictive operations. This allows you to forecast performance degradation and potential failures before they occur, preventing costly downtime and improving asset reliability.

5. The Next Frontier: Multi-Agent Workflows

While individual AI tools are powerful, the future of engineering automation lies in multi-agent systems. Imagine a “digital engineering team” where specialized AI agents collaborate to execute complex, cross-functional tasks without constant human hand-holding.

In a multi-agent workflow, different agents act as specialists—one might be an expert in reading requirements, another in CAD generation, and another in physics simulation. They communicate with each other to complete an objective that spans multiple software platforms.

Synera

Synera, a leading platform for engineering process automation, is pioneering the use of connected AI agents. In the recent AI Engineering Bootcamp (Session 3), Ram Seetharaman (Head of AI at Synera) demonstrated a live multi-agent workflow that orchestrates the entire design-simulate-iterate loop.

In this example, a “Manager Agent” in Synera interprets requirements and delegates tasks. It triggers a “Geometry Agent” to modify a design logic, which then hands off the new geometry to a “Simulation Agent” (powered by SimScale) to validate performance. The results are fed back to the Manager, which decides whether to iterate further or finalize the design.

By chaining these agents together, you eliminate the coordination bottlenecks that often stall projects for days or weeks, allowing for continuous, “always-on” engineering operations.

The Engineer’s Future: AI as an Amplifier

Across all these stages, the theme is the same: AI is an amplifier for engineering expertise. It tackles the repetitive, time-consuming, and data-heavy tasks that slow down innovation. This allows you, the engineer, to focus on what you do best: problem-solving, creativity, and making the critical decisions that lead to breakthrough products. The companies and engineers who embrace these tools today will be the ones shaping the future of the industry.

The post AI Tools for Mechanical Engineers: Transforming Your Workflow appeared first on SimScale.

Eliminating Bottlenecks

The session brought together three distinct perspectives on how to operationalize AI in production environments: Ram Seetharaman (Head of AI, Synera) on Agentic AI, Matthias Bauer (Director of Software Development, Autodesk / Founder, NAVASTO) on Physics AI, and David Heiny (CEO, SimScale) on the cloud-native infrastructure that binds them together.
The consensus? The industry is moving past the “chatbot” phase. We are entering an era where AI Agents orchestrate complex tools to automate busy work, and Physics AI provides instant feedback loops, allowing engineers to traverse design spaces at unprecedented speed.

Key Takeaways:

1. Agentic AI is the “Digital Engineer,” Physics AI is the “Calculator”

The session clarified the distinction between the two critical types of AI. Agentic AI (using LLMs) acts like a digital employee—reasoning, planning, and orchestrating tools to handle complex processes like RFQ responses. Physics AI (using GNNs) acts as an ultra-fast solver, providing instant performance predictions to accelerate the design iterations that the agents (or humans) generate.

2. Integration is the Multiplier (The “Electric Motor” Analogy)

Matthias argued that simply swapping a solver for an AI model isn’t enough. He compared it to the industrial revolution: replacing a steam engine with an electric motor didn’t yield efficiency gains until factories were redesigned around the new power source. Similarly, AI only delivers ROI when deep-integrated into the tools engineers already use (like CAD), rather than sitting in a silo.

3. Trust Comes from Traceability, Not Blind Faith

A major barrier to AI adoption is the “black box” problem. The panel emphasized that trust is built through auditability. For Agentic AI, this means viewing the “chain of thought”—seeing exactly which tools the agent used and why. For Physics AI, it means statistical validation and “traffic light” confidence scores that tell an engineer when a prediction is reliable and when to fall back to traditional simulation.

4. The “Junior Engineer” Model

AThe most practical way to deploy AI today is to treat it as a “junior engineer.” It can autonomously handle tedious tasks (like meshing, setup, or initial design sweeps) and present 80% complete work for expert review. This keeps humans in the loop for critical engineering judgments while removing the bottleneck of manual execution.

Watch the full webinar recording below. And if this seems interesting, be sure to check out the rest of the series!

The post Webinar Highlights: AI-Native Engineering Workflows appeared first on SimScale.

Eliminating Bottlenecks

The discussion broke down the two primary bottlenecks in engineering: simulation lead time (setup) and simulation cycle time (computation), and explored the AI technology that lets you transform engineering processes by effectively eliminating them.

We heard some great insights from Brian Sather from nTop, explaining the importance of a robust geometry pipeline for effective design exploration. Jon Wilde described SimScale’s approach to tackling the bottlenecks in simulation workflows that unlock the full potential of AI-driven engineering.

Key Takeaways:

1. Solving This Needs a Two-Pronged Solution

Physics AI (using GNNs/PINNs) learns physics to deliver instant predictions, crushing the computation bottleneck. Engineering AI (using LLMs) understands user intent to automate and orchestrate entire multi-step processes, crushing the setup and lead time bottleneck.

2. To Scale AI, You Must Solve Data Generation

One of the most significant challenges in scaling AI is assimilating or generating training data. Here, the robustness and speed of geometry generation is key, and traditional CAD models can struggle. We looked at how “computational design” tools can algorithmically generate thousands of valid design variants, creating the synthetic data needed to train a reliable Physics AI model.

3. A Connected Toolchain Is Critical

Eliminating process bottlenecks is only possible with a seamlessly connected toolchain with limited sprawl. The session demonstrated how to build tight, AI-driven optimization loops involving nTop’s implicit models that can be read directly by SimScale, eliminating manual prep work and ensuring a robust transfer from geometry to simulation.

4. AI Agents Are the New “UI” for Democratization

A live demo of SimScale’s Engineering AI agent showed how non-experts can now drive complex simulation much faster. By using natural language (e.g. “optimize this heat sink for me”), a user can trigger an agent to orchestrate CAD, simulation, and optimization in the background. This moves simulation from a specialist-only tool to a capability accessible to the entire organization.

Watch the full webinar recording below. And if this seems interesting, be sure to register for the rest of the series!

The post Webinar Highlights: Scaling AI-Powered Simulation Across Teams appeared first on SimScale.

An AI Masterclass – How to Fit Months into Hours

The highlight of the session was a real-world case study from Armin Narimanzadeh, Senior Thermofluids Expert at Convon (part of HD Hyundai). Armin shared his first-hand experience of using SimScale’s AI-powered simulation to optimize a hydrogen ejector pump, building a reusable Physics AI model that produces instant performance predictions for new designs.

This transformative approach reduced a design optimization process that previously took months down to under an hour, enabling rapid iteration and data-driven decision-making.

The discussion, featuring insights from Mike LaFleche of PTC and Steve Lainé of SimScale, explored the crucial role of a cloud-native ecosystem in making these workflows possible and how to overcome common blockers like data availability and trust in AI.

Key Takeaways:

1. AI is an Amplifier, Not a Replacement for Expertise

A recurring theme was that AI serves as a powerful tool to amplify your engineering expertise. Armin emphasized that while the AI model delivered incredible speed, his engineering expertise was still crucial to guide the optimization, validate the final results against CFD, and make the final design decisions. The goal is to empower experts, not replace them.

2. The “Months to Hours” Transformation is Real

The most powerful takeaway was the quantifiable impact on the product development cycle. Having invested in the initial model training and data generation, Armin’s team now has a reusable AI model that can generate a new, optimized design for their ejector in under an hour. This is a game-changing acceleration that directly impacts business agility.

3. A Cloud-Native Ecosystem was Key

This level of automation and speed is only possible when the entire toolchain is cloud-native. The seamless, API-driven connection between a parametric model in Onshape and the simulation in SimScale was essential for automatically generating and testing hundreds of design variants to firstly map the design space and then to explore and optimize within.

4. You Can Start Now, Even Without Perfect Data

Armin carefully tested different training data sets to find the dataset ‘sweet spot’ – how much data was needed to build an accurate model. He found that the number of samples needed was not as large as originally expected, allowing him to refine his approach for future projects.

Watch the full webinar recording below. And if this seems interesting, be sure to register for the rest of the series!

The post Webinar Highlights: Kickstarting Engineering AI in Manufacturing appeared first on SimScale.

The question is no longer if you should adopt AI in your engineering workflows, but how to do it effectively and at scale. Across industries, engineering leaders are experimenting with AI pilots, but many are finding it difficult to move from isolated experiments to production-scale adoption. The result is often a collection of stalled initiatives and an uncertain return on investment.

For many engineering teams, the core challenge is the lack of a clear blueprint – what does success look like? Unlike other business functions, engineering has unique data and workflow complexities. Your proprietary design data is complex, multimodal, and not easily digestible by the large, general-purpose AI models that have captured the public imagination. A “one-size-fits-all” strategy simply doesn’t work.

A successful AI strategy in engineering isn’t a case of finding one algorithm or system to do everything. The key is to identify points in your workflow where it can add the most value – in other words so you can tackle the most significant bottlenecks in your product development cycle. This is the path from experimentation to transformation.

Two Bottlenecks, Two AIs

The product development process is a race against time. The goal is to shorten the loop from the moment a design is created to the moment its performance is fully understood. This delay is caused by two fundamental bottlenecks:

1. Simulation Lead Time: This is the manual effort and human waiting time. It includes the handoffs between teams, CAD preparation, meshing, and the entire simulation setup process. It’s the time spent preparing to get an answer.
2. Simulation Cycle Time: This is the raw compute time. It’s the hours—or even days—your high-fidelity solvers need to run to deliver a single, accurate result. It’s the time spent waiting for an answer.

To truly accelerate innovation, you need to attack both bottlenecks simultaneously. SimScale does this with a two-pronged AI strategy, deploying a purpose-built AI to solve each problem:

• Engineering AI: An autonomous agent designed to eliminate Simulation Lead Time by automating the entire workflow from setup to execution.
• Physics AI: A predictive system designed to eliminate Simulation Cycle Time by providing instant performance insights without running a full simulation.

Engineering AI in Action: Your New Autonomous Teammate

Consider the following engineering challenge: designing a cold plate for cooling electronics. The goal is to find the right trade-off between thermal performance (cooling) and pressure drop (efficiency).

Usually, an engineer would manually or programmatically set up a simulation for each design iteration—a repetitive and time-consuming process. Engineering AI transforms this. It acts as an autonomous agent that can perceive the simulation context (the geometry, the physics) and execute the entire workflow based on simple instructions. It becomes a new kind of teammate, handling the mundane tasks so your experts can focus on analysis and innovation.

The payoff here isn’t just about speed; it’s about liberating your most valuable engineering talent. By automating the setup, you empower your team to focus on the high-value work of interpreting results and driving design decisions.

Physics AI in Action: Real-Time Insight

Even with a fully automated setup, complex physics can take hours to compute. This inherent delay makes rapid, comprehensive design space exploration impossible.

This is where Physics AI comes in. By training on the results of previous high-fidelity simulations, it learns the physics of your design. It can then infer the performance of a new design variant in seconds, without ever running a full solver. This transforms the workflow from a slow, iterative process into an interactive design session.

As noted in our recent webinar with Ian Pegler from NVIDIA, this capability finally makes comprehensive design space exploration feasible for complex problems. It’s the difference between testing a handful of ideas and exploring thousands.

The Blueprint: How Engineering and Physics AI Work Together

The true revolution happens when these two systems work in concert. They are symbiotic, creating a workflow that is both automated and instantaneous. This is a day in the life of the modern, agent-augmented engineering team:

1. An engineer wishes to kick off a project to optimize a design (let’s say it’s a hydraulic manifold). Instead of digging out CAD files and simulation templates, they instead brief a customized Engineering AI agent in SimScale to perform the study, based on some simple instructions.
2. As part of the study, the AI agent chooses to make use of a relevant pre-trained Physics AI model which can immediately run inferences to provide a real-time prediction of each design’s thermal and hydraulic performance. The engineer, supervising the process, can interactively explore dozens of variations in minutes.
3. The human engineer and Engineering AI discuss the foremost design candidates, selecting a promising variant which demonstrates the desired characteristics spanning performance, cost and manufacturability.
4. Then it is back to the Engineering AI agent to autonomously set up and run a full, high-fidelity simulation in the background for final validation, while the engineer is already moving on to the next creative task.

The total time from a design change to a fully validated performance insight drops from days or hours to minutes. This isn’t a minor improvement; it’s a fundamental acceleration of the entire innovation cycle.

With expert knowledge captured in the AI agent, the whole workflow described above is now accessible to less expert, junior engineers. It is this democratization that allows engineering teams to scale and move faster.

Your Starting Point

Look at your own organization and ask: which bottleneck is slowing you down the most? Is it the manual lead time spent on setup and preparation, or the computational cycle time spent waiting for solvers?

Answering that question will reveal your starting point for a scalable AI strategy that delivers tangible results. If you would like to discuss your strategy with one of our experts, get in touch today.

Catch up on the webinar

Learn more about deploying AI in engineering with SimScale and NVIDIA as part of engineering.com’s Digital Transformation Week

Watch On Demand

The post A Day in the Life of Engineering AI appeared first on SimScale.

AI in engineering is only just starting to break out into the mainstream. We recently surveyed 300 engineering leaders and we found that only 7% had mature AI programs in place, with 42% actively working on pilots, but over half having still not yet started anything serious.

But while we saw widespread acknowledgement that AI in engineering has huge potential, only 3% of companies adopting it are already reaping the significant rewards that are possible. 

Put simply, there is a huge opportunity for engineering organizations to build a competitive advantage by adopting this technology in the right way.

This article cuts through the noise to showcase seven ways that we see AI delivering value to our customers. These are not futuristic concepts; they are tangible solutions that directly address the most time-consuming, error-prone, and knowledge-intensive aspects of engineering.

Simulation Democratization: AI for Every Engineer

Traditionally, physics simulation has been a specialized field, confined to a few experts with access to expensive software and hardware that’s complicated to learn and use. This creates a dual bottleneck: projects stall waiting for expert review, and deep engineering knowledge is difficult to scale across the team.

The AI Solution: The goal of simulation democratization is to put these powerful tools safely into the hands of every engineer. Cloud-native platforms already provide much easier browser-based access, but the real key is Agentic AI acting as a force multiplier for expertise. It captures the proven methods of senior engineers and embeds them into reusable templates and guided workflows. This gives every engineer an AI-powered co-pilot that walks them through complex processes step-by-step, interacting using natural language. Opening up access to simulation in this way ensures consistency and reduces errors, while scaling knowledge across the entire organization.

RFQ and RFP Automation: Winning More Bids

Responding to a Request for Quote (RFQ) or Request for Proposal (RFP) is a high-pressure, make-or-break moment. Engineering teams must quickly assess feasibility, optimize designs, and produce reliable cost estimates under tight deadlines. This often leads to rushed proposals, reduced win rates, and shrinking profit margins.

The AI Solution: AI-driven automation transforms the RFQ process. An Agentic AI can read an RFQ, instantly generate relevant design concepts, and automatically validate them using physics simulation. This allows teams to explore multiple options, identify cost and feasibility risks early, and package proposal-ready reports in hours instead of days.

Accelerating R&D and Product Innovation

Traditional R&D is slow, expensive, and risky. Heavy reliance on costly physical prototyping and limited resources means teams can only explore a handful of design options, leaving breakthrough ideas undiscovered and inflating budgets.

The AI Solution: AI-enhanced simulation tackles R&D challenges on two fronts: cost and speed. It enables virtual prototyping, where thousands of digital experiments replace physical models to cut material waste and lab expenses. Simultaneously, it powers massive design space exploration, where generative workflows create and test thousands of concepts in parallel. This combined approach allows teams to compress development timelines from months to days while making smarter, more cost-effective design choices.

Agentic Workflow Automation: Connecting the Dots

Engineering processes often involve a maze of manual handoffs between different tools and teams for tasks like geometry prep, simulation, and reporting. Each step introduces potential delays and human error, wasting valuable engineering time that could be spent on innovation.

The AI Solution: Agentic workflow automation uses AI to connect these disparate steps into a single, seamless process. AI agents can coordinate everything from CAD input and multi-physics analysis to optimization and reporting, all without manual intervention. By integrating with existing tools like CAD and PLM systems, these automated workflows plug directly into established engineering processes. Check out our recent webinar: The Rise of AI Agents in Engineering – What Can We Expect?.

Real-Time Digital Twins: Bringing Designs to Life

Digital twins are powerful, but traditional versions are difficult to implement, often requiring heavy modeling effort and costly, manual data integration. As a result, the insights they provide often lag behind reality.

The AI Solution: AI is making real-time digital twins a practical reality. The key is using lightweight Physics AI surrogates – AI models trained on high-fidelity simulation data that can run in real-time. By connecting these surrogates to live sensor data from products in the field, engineers can get instant predictive insights into performance, spot potential failures before they happen, and feed real-world data back into the R&D process.

AI in engineering is here – it just needs to get from the lab into the loop

The engineering industry’s gap between AI ambition and execution is a clear signal that a focused approach is needed. The path to closing this gap is not about “doing AI” in the abstract, but about deploying it with purpose. It’s likely that your organization is already pursuing initiatives like those explored in this blog, AI simply redefines what is possible to achieve with them:

The use cases for AI in engineering detailed in this blog—from further democratizing simulation to automating commercial responses—are tangible, real-world applications that directly solve the most significant productivity bottlenecks in the modern engineering simulation process. They are the practical bridge that connects the ambition for AI to its successful execution.

For engineering leaders, the call to action is to shift from tactical experimentation to strategic transformation. This means investing in the open, cloud-native platforms capable of supporting this new generation of embedded, agentic, and collaborative AI. The ultimate goal is not to replace the invaluable expertise of engineers, but to fundamentally enhance it. By creating “machine-in-the-loop” workflows, we can supercharge their creativity, multiply their impact, and free them to focus on what they do best: solving the world’s most complex challenges and engineering the irreplaceable.

Ready to see how AI can transform your engineering workflows?

Get in touch with one of our specialists to learn how you can start applying these use cases today.

The post Real World Engineering Applications of Artificial Intelligence appeared first on SimScale.

For engineering applications of AI, this metaphorical elephant is a tough one to dodge.

It is, of course, data – the lifeblood of any AI model.

Let’s take a step back and understand why.

We’ll see what can be learnt from other AI applications and look at how leading innovators are solving this bottleneck in the engineering space.

Lessons from LLMs: Data is the Determining Factor

The meteoric rise of LLMs like GPT, Gemini, and Claude is often attributed to exponential growth in computing resources and model size. These advances, however, only tell part of the story. These leaps in performance have been accompanied by a corresponding revolution in data engineering.

The extensive pre-training phase of LLM development requires meticulously curating, cleaning, deduplicating, and sequencing trillions of diverse text tokens to maximize learning efficiency and resulting model performance.

In the same way we have to be careful about what we say and how we behave in front of our children, we have to be careful about what makes it into training data sets.

Although the leading AI companies are understandably cagey about exactly how much time and effort (and precisely what) goes into the pre-training phase of model development, the accepted wisdom in the world of data science is that 80% of time and effort goes into data collection and preprocessing, with only 20% spent on modeling. So that’s a lot!

To summarize, the old adage that has been echoing around modeling and simulation teams for decades holds as true today as it ever did:

“Garbage in, garbage out”

Every wise simulation engineer, ever

In engineering simulation, where data is sparse, high-fidelity, and distributed, these lessons are especially relevant. AI systems are only as good as the data they ingest, and they need data quality over quantity.

But what does this mean for physics-based simulations, with their multidimensional numerical fields and complex geometric constraints?

You may, or may not, be sitting on a goldmine of data

All engineering organizations have data, most likely too much data!

Whether that data can be directly used to train useful AI models is another question entirely.

In a recent industry survey, we asked engineering companies about the blockers they encounter when trying to implement AI simulation workflows.

Difficulties in accessing and using data was the top concern raised!

Why is engineering data so often siloed?

Engineering simulation starts with a new design candidate and/or a set of boundary conditions to investigate. It ends with an evaluation of some pre-determined KPIs that tell you how well the design performed with respect to your initial requirements.

Considering the relative sizes of these puzzle pieces, there is a strong argument for discarding the model itself. After all, it can always be re-run if you really need it again… provided you have a good record of what the inputs were.

WARNING: This is a ‘pre-AI’ way of thinking.

The problem with this way of working is that the data that does get systematically retained is not at all AI-ready.

It might be usable for relatively simplistic forms of regression analysis, but it does not usually include the geometric inputs and full field outputs needed to train GNN or PINN models.

The data you really need is now fragmented over many systems, if it still exists at all. Maybe a PLM environment holds the CAD data archive, while KPIs and post processing images might be stored in another database, or maybe even a spreadsheet or folder structure on a shared drive.

Any given simulation model could be languishing on an external storage drive or cluster filesystem, or perhaps it was deleted to save space during a recent de-cluttering exercise. Who knows? It will certainly be a lot of work to find out, gather all the data together, and fill in the blanks.

The reality of legacy tools: a data nightmare

The situation I describe above is often the result when simulation software is used without a data management system in place—something very few teams have (only 5% in a 2019 survey). Legacy tools (whether run on-premises or in the cloud) leave a trail of files and folders behind them that need to be gathered up and archived either manually or using another tool to do the same automatically. Either way, order does not come easily.

Fragmentation of data stifles any attempt to automate data pipelines or scale AI deployment. The time and manpower required just to compile, validate, and align datasets quickly exceeds the cost of actual model training. The result is a cycle – AI pilots thrashed by unresolved data engineering workloads, and promising projects failing to leave the lab.

Cloud-native simulation includes built-in data management

Contrast this with a vertically integrated cloud simulation stack like SimScale. All simulation data generated on the SimScale platform persists (unless deleted) and remains organised and ready for further use or reuse. A single, cloud-native repository aggregates past and current projects, auto-tagged by discipline, physics, and defining parameters. Built-in AI infrastructure then directly taps into this data store, meaning that there are no obstacles to be navigated in between data generation and model training.

SimScale’s Physics AI model training dashboard provides complete control over model publication, versioning and sharing between users and teams. The same framework allows users to access pre-trained foundation models and incorporate them into AI-driven workflows.

What a data goldmine looks like: A real-world example

We recently collaborated with Nantoo, one of our customers designing garden power tools, to see whether the simulation data they had generated over a lengthy product development process could actually be directly used for AI model training. The interesting thing about this particular experiment was that the data had originally been generated with no future intention to use it for AI model training.

In this case we took the simulation runs as they were, without modification, and were able to quickly train a model using SimScale’s built in Physics AI. The results were impressive, and really show how using a cloud-native platform gives you a massive boost when getting AI adoption off the ground.

Why was the process so straightforward? The Nantoo team have been running simulations on SimScale for a while now, so they have built up a large library of runs, and SimScale saves all of your simulation data in the cloud – neatly organised into projects and analyses. They had also followed good practices like consistent model setup and orientation, naming conventions, and the like. 

Since the data was in the cloud and ready to go, all that was needed was for the Nantoo team to share the relevant projects with us (a matter of a just few mouse clicks) and we could immediately create a copy of the project, dive into the data and start model training using SimScale’s built-in AI infrastructure.

Nantoo’s experience illustrates the transformative potential of this technology when deployed in a cloud-native environment. The key: centralized, structured data access and full traceability—capabilities supported by SimScale’s web-native, cloud-based infrastructure. By default, SimScale customers bypass much of the chaotic data wrangling endemic to on-premise tools and are positioned to unlock AI at pace.

Checklist: How to future-proof your data

Now is the moment to audit your simulation data landscape and take strategic action. The companies making the leap to AI-enabled engineering are those who treat data not as an afterthought, but as a first-class product. Take your initial step by evaluating your digital infrastructure—where does your data live, and how quickly can you put it to work? The future of engineering simulation belongs to those who master data, from preprocessing to deployment:

Quality

This is where the biggest parallels to LLM development can be drawn. The quality of a model’s predictions are directly related to the quality of the input data. This should cover many aspects, including mesh quality and model convergence.

Consistency

Because machine learning models are learning patterns between inputs and outputs, that data must be consistently structured, for example using the same flow direction and model orientation, and the same topology and boundary condition types. If you need to predict a specific scalar field, KPI or integral value, that data must already be present in each input run.

Accessibility

Model training requires ready access to the data. If it is scattered and fragmented, this will not only be a huge time sink to gather it together but it will also be very difficult to ensure 1) Quality and 2) Consistency. The simulation, data and model training environments need to be tightly integrated.

Ready to roll? 

Identify one simulation process with accessible, high-integrity data, and commit to piloting your first AI-driven workflow—powered by a cloud-native, AI-native stack. The path to transformative innovation begins with your data.

The post Engineering Data Management – The Key to Transformation appeared first on SimScale.

I no longer think about managing local files, software versions, license servers or finding the right compute resources (enormous laptops, workstations or clusters).

Since all of that complexity is taken care of behind the scenes, complex simulations become as easy to work with, share, and collaborate on as a Google Doc. 

It means that the cloud-native simulation engineer can concentrate on the ‘what’ and not be distracted by the ‘how’.

It’s no surprise then that almost all new entrants to the simulation market in recent years have been cloud-native tools. The benefits I outlined above are compelling arguments for using this architecture to build a next-generation tool. But it goes much deeper than an amazing user experience, with greater productivity and cost-efficiency. Yes – these are all nice-to-have advantages, game-changers even… but the playing field is shifting.

That was then, this is now – the era of AI-native simulation

Today, the pace of change has accelerated and the world of engineering sits on the cusp of a new technological epoch: the age of artificial intelligence. This new era brings a great potential to unshackle us from lots of the slow or manual parts of simulation and analysis. It can gallop through workflows and deliver thousands of predictions in a matter of seconds.

It can gather all the data we need to make decisions, so we can focus on engineering and value-add tasks.

AI can do all of these things in an instant if, and only if, we remove all obstacles from its path. That means no shuffling data from one place to another, translating, exporting and importing, uploading or downloading. The whole technology stack, physics solvers, AI solvers and data all need to be co-located and seamlessly integrated.

Ask the audience, what’s stopping you?

In a recent survey we conducted across 300 engineering leaders, we asked what were the most significant barriers to AI adoption in their simulation teams. The responses show that data accessibility and infrastructure are seen as two of the most critical factors.

Companies that are locked into on-premises tool stacks are faced with a good deal of complexity when ‘retrofitting’ AI into their existing processes and workflows, perhaps requiring them to develop or hire specialized expertise to do so.

What these results demonstrate is that it is infrastructure and platform that are the biggest barriers to adoption, and not resistance to change, lack of willingness or budget.

Watch out: cloud-native adopters are pulling ahead

Unsurprisingly, we found that nearly every company contacted by the survey was either actively working on incorporating AI into their simulation programs or was planning to do so in 2025. Even so, only 7% of respondents were already at the stage of having mature AI programs in place. 

What is special about those 7% though?

This breakdown very clearly show that the likelihood of a company being able to successfully leverage AI is closely linked to the type of technology stack they are using.

Organizations using cloud-native simulation tools are 3x more likely to have mature AI programs and 6x more likely to have clean, centralized data—critical for scaling AI. They are also twice as confident in achieving AI goals within the next 12 months.

The question is no longer, “Should we go cloud-native?” but, rather, “What is at stake if we don’t?”

We are still at the start of the AI revolution in engineering. The rate of development is fast, and while the engineering community has seen enough evidence to convince them that significant benefits can be achieved, there remains a big gap between expectation and execution. The question is how to progress from pilots and proof-of-concept to profitability?

Survey results show that for every organization seeing significant benefits of AI in production, there are ten others who see the same opportunity but have not yet been able to realize the gains.

Survey respondents also make it clear that legacy tools are not simply slower or less convenient—they are a source of active risk. Two-thirds of all respondents state it is difficult to integrate AI practices using their current stacks. These difficulties are multiplied in on-premise setups, which tend to fragment data and create manual, brittle integrations between systems.

Cloud-native, AI-native engineering simulation: build your AI powered future on the right foundations

Cloud-native organizations are more than just early adopters; they are realizing AI’s potential at a greater rate and, crucially, preparing for scale.

But why do cloud-native tools have the advantage?

I can think of three primary reasons:

1. AI needs data, and quickly. Transformational AI cannot be a ‘bolt-on’ capability. It needs a deep and immediate connection to simulation models and data. And that data needs to be unified and structured. The cloud is ideally suited to creating this sort of environment.
2. Once you have built out AI-powered simulation workflows, some of the most significant benefits will come from being able to further democratize access and foster collaboration between engineers and teams, or even between AI agents operating across your toolchain. Again, cloud-native platforms like SimScale are designed from the ground up to maximize accessibility and collaboration.
3. Developments in this space, both in terms of software and hardware, are moving extremely fast. Cloud-native, AI-native platforms which are intuitive to use and keep the complexity transparent to the user are the only way to keep your engineering teams at the cutting edge.

Looking to the future, our survey found cloud-native adopters are more than twice as likely to express confidence in meeting AI goals in the next twelve months than their on-premises counterparts.

For those still waiting, the risk is clear. In today’s market, failing to bring cloud-native platforms into your engineering organization’s stack is not just a missed opportunity for efficiency. It is a competitive liability, one that will become more pronounced as AI matures and industry leaders start to benefit from the transformational engineering velocity that it will unlock.

If you are responsible for the future direction of engineering in your company, now is the time to ensure your foundation is ready. The evidence is conclusive: cloud-native software was a competitive advantage. Now, in the era of AI, it is the essential condition for modern engineering success. The next wave of engineering innovation demands it.

Don’t wait for the gap to widen. Close it now and lead your industry into the future.

Learn more: https://www.simscale.com/webinars-workshops/engineering-teams-struggling-realize-ai-opportunity-how-fix-it/

The post AI makes cloud-native software essential for engineering simulation appeared first on SimScale.

An agentic Engineering AI is in active development at SimScale, integrated directly into our simulation platform, and we have customers using it today in alpha testing. This agent is set to make simulation even more accessible, intelligent, and productive, regardless of your experience level.

In this article, we dive into five agentic workflows that we see Engineering AI unlocking, either today or in the near future. The technology is maturing fast and we are excited about what’s coming next!

1. Helping Novice Users Get to a Working Simulation Sooner

Making simulation approachable for new users has always been at the core of SimScale’s mission. Engineering AI agents turbocharge this democratization. Rather than relying on static documentation or guides, an agent understands your CAD model and context in real time. It can proactively guide you step-by-step toward a viable setup—automatically diagnosing missing inputs, suggesting best practices, and flagging potential challenges long before you hit “run.” This transforms onboarding from a knowledge hurdle into a guided, confidence-building experience.

2. Accelerating Model Setup With Intelligent Automation

Setting up a simulation is often time-consuming, particularly when it comes to assigning boundary conditions, materials, or physics models. SimScale’s Engineering AI leverages both geometric information and vast simulation knowledge to streamline this process. The agent recognizes context from your geometry and model setup, recommends or auto-applies suitable settings, and raises issues only when human judgment is needed. The result: faster, more consistent model preparation with less context switching or uncertainty—bringing value to novices and seasoned analysts alike.

3. Guiding the Application of Company Best Practices

Scaling simulation quality and consistency across a distributed team remains an industry pain point. Engineering AI can actively reinforce organizational best practices. It serves as a digital mentor, surfacing internal guidelines and guarding against common pitfalls as users progress through a workflow. Over time, such agents learn from recurring mistakes or successes, driving incremental improvement and consolidating hard-won expertise—no matter the user’s prior experience.

4. Collaborating With Other Agents

Some of the most exciting potential of Engineering AI agents emerges through collaboration—between agents, not just humans. Multiple specialized agents, each with domain-specific reasoning, can work together across complex engineering challenges. Recently, during a joint webinar with Generative Engineering, we demonstrated a proof-of-concept agent-to-agent workflow. Here, SimScale and Generative Engineering agents jointly orchestrated a design space exploration, translating broad goals into practical analyses and autonomously iterating on concepts in the background. This kind of digital teamwork could transform the status quo of tool interoperability and ultimately deliver seamless, agent-powered multi-tool engineering workflows in the near future.

5. Automating RFQ Responses

One high-impact future application of the workflow demonstrated above is automating RFQ (Request for Quotation) responses. Mature engineering agents will soon be able to ingest customer requirements, map them to relevant design specs, set up and run appropriate simulations, and validate that the proposal meets performance criteria—all automatically. For organizations handling large volumes of RFQs or routine design studies, this unlocks significant value: reducing manual workload, speeding up response times, and freeing engineers to focus on high-impact, creative tasks.

Webinar: Is Your Organization Ready for AI?

Looking Ahead

Engineering AI agents represent a step change for simulation-driven design. They reduce ramp-up times, accelerate core workflows, promote best practices, enable multi-agent collaboration, and automate entire processes. At SimScale, we’re building these capabilities directly into our platform so that more organizations and engineering teams can benefit from agentic workflows.

If you’re interested in what AI-accelerated engineering could do for your team, get in touch with us below and be part of this new era of digital engineering.

The post What can Engineering AI do for you? Here are 5 agentic workflows set to transform productivity appeared first on SimScale.