Project
Generative Assembly Design
Rethinking mechanical system design through intelligent, AI-driven workflows
Generative Assembly Design (GAD) explores the future of mechanical assembly design through AI-driven tools that move beyond traditional CAD workflows toward a more integrated, behavior-driven paradigm.
This research focuses on early-stage mechanical systems—such as gears, shafts, and beams—to develop methods that unify conceptual and detailed design within a single continuous process.
A Unified Design Workflow
GAD connects design requirements, functional behaviors, modular component building blocks, and generated geometry into a cohesive workflow. Rather than treating these elements as separate steps, the system integrates them into a dynamic, iterative process.
By combining:
- Interactive geometry generation
- Intelligent component sizing
- Agentic AI assistance
GAD enables a collaborative design experience where the system works alongside designers to iteratively develop and optimize assemblies.
“We see the future of design working through graph-based system interactions powered by geometry generation tools and supported by assembly behaviour simulation. It challenges the traditional thinking about CAD and assembly design.”
Nigel Morris
Sr. Manager, Principal Research Scientist
Understanding Assembly Complexity
A key focus of the research is building detailed assembly examples across multiple engineering domains. Through these examples, the team investigates:
- The complexity of individual components
- The relationships and dependencies between components
- The information embedded within components and their connections
- How assemblies interact with external systems and environments
This approach helps reveal how mechanical assemblies function as interconnected systems, informing the development of more intelligent and adaptive design tools.
Related Resources
2026
Building Academic Connections and Advancing Simulation Research at WCCM/ECCOMAS 2026Autodesk Researchers attended WCCM/ECCOMAS 2026 conference to share…
2025
From Design and Break to Design and MakeExplore the benefits of designing and manufacturing real parts when…
2025
From Design and Break to Design and Make, Part TwoLearn more about the benefits of designing and manufacturing real…
2025
A Domain Adaptation of Large Language Models for Classifying Mechanical Assembly ComponentsThis research evaluates the feasibility of using LLMs to classify…
2025
An Interface-Preserving Level Set Update Strategy for Topology Optimization of Mechanical AssembliesEnables generative design for mechanical assemblies, allowing…
Get in touch
Something pique your interest? Get in touch if you’d like to learn more about Autodesk Research, our projects, people, and potential collaboration opportunities.
Contact us