Author(s): Dr. Axel Richter

Originally published on Towards AI.

The Future of Engineering Design Is Agentic

The next CAD revolution is not a better text box. It begins when an AI system understands a design goal, independently selects the right tools, changes the native model, verifies the result, and uses failures to determine its next step. This closed loop is what turns a copilot into a design agent.

The article explains that “agentic” CAD is more than generative design or a chat interface: it is an executable closed loop that perceives, plans, acts, and verifies against an engineering specification. It outlines the supporting architecture—goal-to-spec workflows, typed tool surfaces (e.g., via MCP), a stateful harness with permissions and transactions, and deterministic verification outside the language model—so agents know when to stop, when to ask humans, and how to remain auditable and reversible. It then compares how this pattern maps to MCAD, ECAD, PCB, and EDA/chip design, noting which ecosystems are making progress (often through partial copilots or scripting) and where the key missing pieces remain: stable native object models, structured intent and context, verifiable evidence, and an organization-ready agent platform. Finally, it highlights common failure modes from “autonomy theater” to weak success metrics, and concludes with practical guidance for starting with narrow, frequent tasks, building read/verify tooling first, and measuring rework and safety—not just speed or demo quality.

Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI