Cognitive Design Systems (CDS), developer of Cognitive Design, the AI-enabled concept exploration platform for mechanical and thermo-mechanical parts, announced a breakthrough research demonstration: a fully autonomous part optimisation workflow powered by their on-premises MCP (Model Context Protocol) server connected to Anthropic’s Claude LLM (link). For the first time, structural components can be taken from raw CAD input to manufacturing-ready, weight-optimized geometry through natural language alone, with every simulation, FEA run, and manufacturability check executed inside a secured, local environment, on deterministic solvers that produce identical, fully traceable results for identical inputs.
A Fully AI-Orchestrated Workflow
Applied to a multi-axis structural component with complex loading conditions and multiple functional interfaces, the workflow unfolds as follows. An engineer directs the system to a CAD file on the local network and specifies the load cases, material constraints, and performance targets. From there, the system identifies which surfaces are functionally untouchable and which regions can be redesigned, then runs a closed optimization loop: generative design, meshing, finite element analysis, stress evaluation, manufacturing-driven design, geometry refinement, repeat. Every iteration is executed autonomously, converging on a feasible structure where material exists only where forces travel. The engineer decides how much to intervene: the system can run end to end without interruption, or the engineer can engage the AI at any stage to refine targets or redirect the search.
Outputs include manufacturing-ready geometry and automatically generated dashboards covering mass-vs-iteration trajectories, stress convergence plots, and safety factor validation. Every decision step is logged. Every result is fully traceable, grounded in physics-driven algorithms rather than statistical inference, and backed by a data trail that meets the documentation requirements of regulated industries.
What used to take weeks of back-and-forth between CAD, simulation, and engineering review now happens in a single uninterrupted one-hour workflow.
What Changes for Your Engineering Teams
“Engineers have always known what great design looks like. The bottleneck has never been judgment; it has been execution time. This removes the bottleneck, while keeping the engineer in the loop at every step.” — Rhushik, CEO, Cognitive Design Systems
The numbers speak directly to program economics. Where a typical development cycle yields 3 to 5 explored concepts before timeline pressure forces a decision, this workflow produces 50+ fully analyzed variants in the same period. Engineers stop operating software and start making decisions: which trade-off best fits the application, which geometry holds up under the edge-case load, which configuration hits the weight and cost targets simultaneously.
That shift, from tool operation to engineering judgment, is where competitive advantage is built. Underpinning it is a deterministic computational core: same input, same output, every time. The AI suggests and orchestrates; the engineering solvers build, validate, and score. Engineers get intelligent automation without sacrificing the repeatability and traceability their certification workflows demand.
Why On-Premises Changes Everything
Most AI tools for engineering require sending your design data to a cloud server. For companies working on defense programs, space hardware, or proprietary automotive platforms, that is not an option. ITAR, EAR, and equivalent regulations make cloud-routed design data a compliance problem, not just a security preference.
Cognitive Design’s MCP Server runs on your own infrastructure. Your CAD files, your simulation results, your design logic: none of it leaves your network. The AI operates with full capability inside the customer’s security perimeter. Regulated industries get the same performance as everyone else, without the legal exposure.
A Note on What This Is
This demonstration reflects a beta-stage capability, actively being refined toward a production-grade release. CDS is using it to validate the autonomous workflow architecture and shape the next version of Cognitive Design. The final product will be a refined, production-grade implementation of what this prototype proves is possible.
CDS is actively looking for Beta partners in aerospace, defense, space, and automotive to test these capabilities on real components and real constraints. If your organization designs structural or thermo-mechanical parts and wants to explore what this could mean for your workflow, CDS welcomes the conversation.
Read more about AI and CDS’ MCP Server for engineering here.
Manufacturing & Engineering Magazine | The Home of Manufacturing Industry News
