IndustryForge-27B: A Domain-Enhanced Multimodal Foundation Model for Industrial CAD

2026-07-30

AI Summary

This work addresses the complex demands of industrial CAD scenarios—such as engineering drawing comprehension, parametric modeling script generation, Windows COM API invocation, and end-to-end support for part-to-assembly workflows—which generic multimodal models struggle to fulfill. Building upon Qwen3.5-VL-27B, we integrate six domain-specific industrial CAD corpora (comprising approximately 52,000 samples) and employ a unified multi-task supervised fine-tuning strategy to develop the first domain-enhanced multimodal foundation model that jointly supports visual understanding, code generation, and COM interface operations for mainstream CAD software. The resulting model achieves an average improvement of 33.65 percentage points across four CAD-specific benchmarks, substantially outperforming GPT-5.4, while maintaining stable or slightly enhanced performance (+1.56 pp) on eleven general capability benchmarks without catastrophic forgetting, establishing it as a robust foundation for industrial AI agents.

Abstract

Automating industrial CAD design and manufacturing places distinctive demands on multimodal foundation models: the model must see engineering drawings and 3D geometry screenshots, write correct parametric-modelling scripts and Windows COM API code, and cover the full range from single parts to assemblies. General-purpose multimodal models fall short on these tasks, while single-task fine-tuning is too narrow to support the diverse calls that upper-layer agents issue. We build IndustryForge-27B on top of Qwen3.5-VL-27B by curating and integrating six industrial-CAD sub-corpora totalling $\sim$52k multimodal samples---CAD Visual QA (CAD-VQA), parametric CAD code (text2cadquery), assembly-level CAD code (text2cadquery-assembly), and three COM sub-corpora for Inventor / SolidWorks (com_2d / com_3d / com_assembly)---and training with a unified multi-task SFT recipe. Across four CAD-domain benchmarks IndustryForge-27B lifts the base model by $+33.65$~pp on average and outperforms the strong closed-source model GPT-5.4 on all four; across eleven general-capability benchmarks it retains, and slightly improves upon, the base model ($+1.56$~pp mean, no catastrophic forgetting). IndustryForge-27B will serve as the common substrate for downstream industrial-agent projects, providing a unified starting point for a full-stack industrial agent that spans from CAD design to industrial-software operation, from parts to assemblies, and from single-shot generation to closed-loop self-improvement.

Problem

• industrial CAD
• multimodal foundation model
• parametric modelling
• COM API
• assembly design

Innovation

• multimodal foundation model
• industrial CAD
• parametric modeling
• COM API
• multi-task SFT