Introduction: A New Era of Engineering

Artificial intelligence is moving beyond chatbots, image generators, and general-purpose digital assistants. In engineering, AI is beginning to influence how products are conceived, designed, tested, optimized, and prepared for manufacturing.

Industrial AI represents this next stage of digital transformation. Rather than generating content alone, it combines artificial intelligence with engineering knowledge, scientific principles, simulation technologies, and industrial data to support decisions that affect real-world products and systems.

Dassault Systèmes is pursuing this vision through its CATIA design and engineering portfolio, the 3DEXPERIENCE platform, and its Virtual Twin technology. The objective is to connect human expertise with intelligent digital capabilities throughout the product development lifecycle.

For engineers, CAD designers, simulation specialists, and manufacturing organizations, this approach could change not only how quickly products are developed, but also how engineering knowledge is applied and reused.

What Is Industrial AI?

Industrial AI is the application of artificial intelligence to engineering, manufacturing, scientific research, infrastructure, and other complex industrial environments.

Traditional AI applications can generate text, recognize patterns, or produce images. Industrial applications face additional requirements: their outputs must be relevant to physical systems, engineering constraints, safety considerations, and manufacturing requirements.

For example, generating a visually convincing mechanical component is not enough. The component may also need to satisfy dimensional constraints, withstand specified loads, fit into an existing assembly, and remain feasible to manufacture.

Industrial AI aims to connect intelligent automation with this engineering context.

Key capabilities include:

• Engineering knowledge integration: Using established engineering rules, product information, and organizational expertise.
• Design automation: Accelerating repetitive modeling and product development activities.
• Simulation-assisted decision-making: Evaluating design alternatives using modeling and simulation workflows.
• Knowledge reuse: Making existing engineering information easier to find, understand, and apply.
• Lifecycle collaboration: Connecting design, systems engineering, simulation, and manufacturing within a shared digital environment.

The distinction is important. Industrial AI is not simply about adding an AI chatbot to CAD software. It is about integrating AI into engineering workflows where context, verification, and traceability matter.

Why Is Dassault Systèmes Focusing on Industrial AI?

Modern products are increasingly multidisciplinary. An electric vehicle, for instance, combines mechanical structures, electrical systems, embedded software, thermal management, sensors, and safety requirements.

These interconnected systems create significant challenges for engineering teams. A change to one component can affect several other parts of the product, requiring additional analysis, validation, and coordination.

Dassault Systèmes is addressing this complexity by connecting CATIA design capabilities with simulation, systems engineering, product data, and Virtual Twin technologies through the 3DEXPERIENCE ecosystem.

The broader strategy is to move from isolated digital tools toward a connected engineering environment in which people and AI-powered systems can collaborate using a common product context.

This approach is particularly relevant to industries such as automotive, aerospace, industrial equipment, high technology, and advanced manufacturing, where product complexity and development costs are substantial.

CATIA and the Evolution of AI-Powered Engineering

CATIA has a long history in 3D design, surface modeling, mechanical engineering, and complex product development. Its role is expanding as AI becomes more closely integrated with design and engineering processes.

The goal is to help engineers spend less time navigating software, searching for information, and performing repetitive operations—and more time evaluating alternatives, solving technical problems, and improving product performance.

Three concepts are particularly important to this evolution:

1. Virtual Companions: AI-powered collaborators that support engineering, scientific, and business activities.
2. Generative Experiences: AI-driven capabilities that help transform inputs and engineering intent into structured design or engineering outputs.
3. Virtual Twins: Digital representations of products and systems that provide context for design, simulation, analysis, and optimization.

Together, these concepts form an approach in which AI can assist with more than individual commands. It can help coordinate tasks and connect engineering information across different stages of product development.

Virtual Companions: More Than Traditional AI Assistants

One of the most notable elements of Dassault Systèmes’ Industrial AI strategy is its introduction of Virtual Companions on the 3DEXPERIENCE platform.

Unlike a conventional assistant that primarily answers questions, these AI-powered companions are designed to work within industrial workflows, use relevant organizational knowledge, and support multi-step activities.

The platform’s announced Virtual Companions include AURA, LEO, and MARIE.

AURA — The Business Expert

AURA focuses on business knowledge, organizational information, and enterprise know-how.

In an engineering organization, important information is often distributed across documents, projects, processes, and teams. Finding the right information at the right time can be difficult, particularly in large organizations with established products and extensive engineering histories.

AURA is designed to help users discover and apply this knowledge more effectively.

Potential applications include finding relevant project information, reusing established practices, and helping teams navigate organizational knowledge.

LEO — The Engineering Expert

LEO focuses on engineering activities and technical problem-solving.

Its intended role is to assist engineering teams with understanding existing designs, navigating product structures, and carrying out engineering tasks through connected CATIA and SIMULIA workflows.

For example, an engineer working on a mechanical assembly might need to understand how components relate to one another, identify relevant design information, or evaluate possible modifications.

An AI companion operating within the engineering environment could help coordinate these activities while using the available product context.

The intended benefit is not to remove the engineer from the process. Instead, it is to extend engineering capacity and reduce the effort required for routine or information-intensive work.

MARIE — The Scientific Expert

MARIE focuses on scientific knowledge and research-oriented activities.

Scientific and engineering problems frequently require knowledge that extends beyond geometric design, including material behavior, physical principles, and complex relationships between variables.

A scientific companion can help users access relevant expertise and support research workflows in which scientific understanding is essential.

Its role complements the broader Industrial AI vision by connecting scientific reasoning with engineering and industrial applications.

Availability and supported capabilities can vary by platform release, subscription, and deployment. Consult the official product documentation for current details.

Generative Experiences: Turning Engineering Intent into Action

Another important element of the CATIA Industrial AI approach is Generative Experiences.

In conventional workflows, engineers often translate design intent into a series of individual operations. They create geometry, define relationships, configure assemblies, document requirements, and check whether the resulting design satisfies the intended constraints.

Generative Experiences aim to automate parts of these processes by converting different forms of input into structured engineering outputs.

Depending on the available capabilities, this can include:

• Generating editable mechanical parts from multimodal inputs.
• Creating assemblies based on predicted mechanical interfaces.
• Generating engineering rules from existing knowledge.
• Transforming engineering documentation into structured systems models.
• Capturing engineering knowledge in reusable templates.

Consider an engineer who needs to create a component from a combination of written requirements, reference images, and existing technical information.

An AI-supported workflow could help interpret the input, propose a suitable model, and generate an editable starting point. The engineer would then review the geometry, verify dimensions, evaluate constraints, and confirm that the result meets the intended requirements.

This distinction is essential: generating a model is only one stage of engineering. Validation, manufacturability, performance, and compliance remain critical.

Virtual Twins: The Engineering Context Behind Industrial AI

AI can be powerful, but its usefulness in engineering depends heavily on the quality and relevance of the information available to it.

This is where Virtual Twin technology becomes important.

A Virtual Twin is a digital representation of a physical product, system, or process. Depending on the application, it can incorporate 3D geometry, engineering properties, simulation results, system relationships, and other relevant information.

Within the 3DEXPERIENCE environment, Virtual Twins can help connect design decisions with engineering analysis and product lifecycle information.

How Virtual Twins Support AI

Imagine developing a new electric vehicle component.

A conventional design process might involve creating the geometry, preparing simulation models, reviewing performance, and modifying the design when issues are discovered.

An AI-supported Virtual Twin workflow could help engineers explore alternatives, identify relevant information, coordinate simulation activities, and compare possible solutions earlier in the development process.

The workflow might include:

1. Creating or modifying the initial 3D design.
2. Connecting the model with relevant engineering requirements.
3. Evaluating alternative configurations using appropriate simulations.
4. Reviewing performance, material, and manufacturing constraints.
5. Refining the design based on engineering results.
6. Maintaining traceability between requirements, decisions, and validated outputs.

The potential benefit is a more connected process in which design and validation inform one another continuously.

Virtual Twins do not automatically guarantee that a product will perform as expected. Their value depends on model quality, appropriate assumptions, reliable data, and suitable validation methods.

Industry World Models: Connecting AI with Physical Reality

A further concept in Dassault Systèmes’ Industrial AI strategy is the Industry World Model.

General-purpose AI systems learn patterns from data. Industrial engineering, however, requires reasoning that is grounded in the behavior of physical systems and the constraints of specific industries.

Industry World Models are intended to provide this context by combining industrial knowledge, engineering relationships, scientific principles, and Virtual Twin information.

This approach is particularly important when the AI is supporting decisions that affect product performance or manufacturing feasibility.

For example, an AI system proposing changes to a mechanical component should consider more than its visual appearance. Relevant factors may include loading conditions, interfaces with neighboring components, material properties, manufacturing processes, and applicable requirements.

Combining AI with engineering models and simulation can help make these recommendations more useful and easier to evaluate.

However, the reliability of a recommendation still depends on the quality of the underlying data, the validity of the models, and the verification process.

Industrial AI and MODSIM: Bringing Design and Simulation Closer Together

Modeling and simulation are fundamental to modern product development.

Traditionally, design and simulation activities can involve separate tools, data preparation steps, and specialist workflows. This can introduce delays when design changes require models to be updated and analyses to be repeated.

Dassault Systèmes’ MODSIM approach brings modeling and simulation closer together within its engineering ecosystem.

When combined with Industrial AI, this creates opportunities to explore design alternatives and evaluate performance more efficiently.

For example, a designer developing a lightweight bracket could use an AI-supported workflow to explore different geometries. Appropriate simulation tools could then help assess stress distribution, deformation, or other relevant performance criteria.

The engineer can compare alternatives and make informed decisions based on the results.

AI may accelerate exploration, while physics-based simulation provides a means of evaluating whether proposed solutions satisfy the required conditions.

This combination is particularly valuable when several competing objectives must be balanced, such as weight, stiffness, cost, manufacturability, and durability.

Practical Applications Across Industries

Industrial AI is relevant to a wide range of engineering disciplines.

Automotive Engineering

Automotive development involves mechanical design, electrical systems, software, safety requirements, and manufacturing constraints.

AI-supported engineering could help teams navigate complex product structures, explore design alternatives, and connect component changes with relevant validation activities.

Virtual Twins can also support the evaluation of product configurations before physical prototypes are built.

Aerospace and Defense

Aerospace programs involve complex assemblies, stringent requirements, and extensive validation processes.

Industrial AI could assist with engineering knowledge reuse, requirements management, design exploration, and the coordination of technical information.

Human oversight and rigorous verification remain essential because aerospace products must satisfy demanding safety and certification requirements.

Industrial Machinery

Manufacturers of machinery and industrial equipment frequently develop customized products with many interacting components.

AI-supported design and assembly workflows could help reduce repetitive modeling activities and improve the reuse of existing engineering solutions.

This can be particularly useful when product families share common components but require customer-specific configurations.

Product Design and Manufacturing

Industrial designers and manufacturing engineers must balance appearance, performance, cost, and production feasibility.

By connecting design tools with engineering knowledge and simulation, Industrial AI could help teams identify potential problems earlier and explore alternatives more efficiently.

Systems Engineering

Modern products increasingly combine mechanical, electrical, electronic, and software components.

AI-supported systems engineering could help organize requirements, establish relationships between system elements, and maintain consistency as designs evolve.

This is especially relevant to software-defined products, where changes in software behavior can influence hardware architecture and overall system performance.

Will Industrial AI Replace Engineers?

The more useful question is how AI will change