CAMBRIDGE, MA, May 28, 2026 – JuliaHub released Dyad 3.0 to bring autonomous simulation agents into the design and validation of complex physical systems. Dyad 3.0 lets engineers use requirements documents, prior designs and test data to generate candidate models, run physics-based simulations, enforce physical and safety constraints, and produce validated models and control code. Dyad is in production with Fortune 100 customers and is being demonstrated in a global livestream.
“AI has transformed software development through agents that combine LLMs with open-source compilers, but engineering physical systems requires the combination of LLMs with a physics compiler that grounds hardware designs in physical laws,” said Dr. Viral B. Shah, CEO and co-founder of JuliaHub. “Dyad 3.0 brings agentic AI directly into the engineering workflow by combining autonomous agents, a multi-physics compiler, high-fidelity simulation, SciML, and enterprise deployment capabilities into one seamless environment. It gives engineers the leverage of AI while preserving the rigor, safety, and verification that physical systems demand.”
Simulation Agents for Engineering Teams
Engineering teams designing aircraft, EVs, semiconductors, utilities, HVAC systems, medical devices and other industrial systems rely on validated models that take time to build. Dyad 3.0 is designed to reduce that manual work while keeping engineers responsible for design direction and final approval.
With Dyad 3.0, engineers can provide a requirements document, a prior-generation design, historical test data and a plain-language request. Dyad agents can assemble a model, explore variations, enforce physical and safety constraints, describe trade-offs and generate validated code for hardware deployment.
The engineer remains the decision-maker – setting direction, reviewing trade-offs and approving final designs. Dyad handles repetitive work such as model construction, controller tuning, simulation execution and toolchain integration.
Engineering AI Gap
AI tools have gained use in software development, but physical engineering faces different requirements. Aircraft, EVs, semiconductors and other industrial systems must be checked against physics, safety and verification limits before teams can trust the design.
General-purpose language models can support analysis or documentation, but they do not validate how physical systems behave under constraints. Dyad combines autonomous agents, physics-based simulation and Scientific Machine Learning (SciML) to work from requirements, simulation models, operational data and prior designs while applying physics constraints throughout the process.
What Is New in Dyad 3.0
- Agentic model generation and design exploration: Agents interpret requirements, propose design candidates, run simulations, and refine models.
- Digital twin workflows for predictive maintenance: Dyad 3.0 supports the design and testing of industrial predictive maintenance applications.
- Agent-driven HVAC system design: The release adds modeling tools, refrigerant splines, expanded library coverage and templates for common system architectures.
- FMU interoperability: Functional Mock-up Unit(FMU) updates improve integration with the broader engineering toolchain.
- Multibody dynamics preview: A preview expands Dyad toward robotics, vehicle dynamics, aerospace mechanisms, and other complex motion systems through 2026.
- Enterprise deployment readiness: Improved installation, configuration, security, compliance, and lifecycle management for regulated and distributed engineering organizations.
Business Impact for Engineering Leaders
Dyad 3.0 addresses cost, delivery timelines, risk control and design exploration in engineering programs.
- Cost: Reduces manual model construction and iteration, lowering engineering hours per program and reducing late-stage prototype rework.
- Revenue: Shortens validated design cycles and help teams work on more programs with the same headcount and reduce time to market.
- Risk mitigation: Physics-based simulation supports design exploration. Safety, regulatory and operating constraints can be encoded and enforced across workflows.
- Innovation speed: Teams can explore larger design spaces, multiphysics couplings and what-if scenarios that are difficult to staff manually.
Proof from the Field
JuliaHub highlighted customer and partner work during the Dyad 3.0 launch event, showing how agent-based simulation is being applied across industrial and regulated engineering environments.
- Commercial HVAC Manufacturer: Uses Dyad’s agentic HVAC library to compress a multi-month system design cycle into a single sprint.
- AI-enabled Aerospace: Uses Dyad as part of infrastructure for predictive maintenance, pilot training, generative design and certification-grade simulation.
- Flight Vehicle Design From Specification: Shows Dyad agents assembling, simulating, and validating NASA’s HL-20 lifting body from a PDF specification.
- JuliaHub Partnerships: Shows how Dyad Is being integrated into partner solutions for hybrid digital twins across industrial applications.
- Digital Twin for Predictive Asset Maintenance: Built a SciML-powered digital twin that predicts pump faults in water distribution networks with over 90% accuracy from four sensor inputs.
- Healthcare Applications: Shows how Dyad 3.0’s installation, configuration and lifecycle management updates support rollout across regulated engineering teams.
Agentic Simulation
Dyad 3.0 combines autonomous agents, simulation, SciML, and enterprise deployment tools. The software is designed that need AI-assisted workflows while validating physical behavior through simulation.
Software agents can support tasks such as drafting, coding or analysis, but physical engineering requires simulation, verification and safety checks. Traditional simulation tools provide engineering depth, but they were not built around natural-language, agent-driven workflows.
Availability
Dyad 3.0 is available from JuliaHub. Engineering teams can view the launch demonstration, review featured customer stories, or request an enterprise evaluation.
Source: JuliaHub
About JuliaHub
JuliaHub, founded in 2015 and headquartered in Cambridge, MA, provides cloud software and tools for scientific and technical computing. The company supports pharmaceutical, aerospace, automotive, electronics and manufacturing customers that use modeling, simulation and computational analysis in their research and engineering work. JuliaHub develops and maintains a platform for building and running applications in the Julia programming language and offers tools for Scientific Machine Learning, digital-twin modeling, circuit simulation and drug-development workflows. The company reports more than 10,000 global users and employs about 110 people worldwide. Its stated mission is to support organizations addressing scientific and engineering problems by delivering secure computing environments and mathematical and machine-learning capabilities.