Driving Automotive Design Transformation: IAT Share Insights in 2025 China Industrial AI Conference

Source: Author: Date:2025-12-18 Read:109

The 2025 China Industrial AI Conference (IAIC 2025), hosted by the China Association for Mechatronics Technology and Application, opened with great fanfare in Hangzhou. 

Centered on the theme "AI Empowerment, Industrial Renewal," the conference convened representatives from government agencies, research institutes, and leading enterprises to explore cutting-edge pathways for integrating industrial AI technology with industrial applications. Liu Yabin, Vice President of IAT Automobile Technology Co., Ltd. and Head of AI.X Lab, was invited to the event. Speaking as a practitioner in "AI-Empowered Automotive R&D Design and Intelligent Manufacturing," he presented a systematic overview of IAT's strategic layout and technological achievements in the field of industrial AI.

Against the backdrop of the global industry's profound digital transformation, industrial AI is emerging as the core driver for high-quality development in manufacturing. As China's leading independent automotive engineering and design solution provider, IAT has been comprehensively advancing the integrated application of AI technology across the entire vehicle R&D process since 2023. 

The company actively explores organic synergy between knowledge engineering, design data, and AI models, building an AI-native digital-intelligent automotive design system centered on the "Automotive User Value - Attribute Large Model" to drive overall R&D efficiency. Currently, IAT has completed the architectural planning for its "Digital-Intelligent Vehicle R&D System" and has successfully developed multiple AI models and intelligent agent tools for automotive design, with several applications already deployed for validation in collaborative projects with automakers.

In the realm of styling design, IAT launched the "TAI" AI Creative Drawing tool. Leveraging text-to-image and image-to-image large model technologies, it assists designers in efficiently generating high-quality interior and exterior styling proposals, significantly shortening the design cycle. This product has now entered the commercialization phase and is being progressively applied in industry practice.

In engineering design, IAT launched a CAD Intelligent Modeling Assistant based on multimodal large models trained on the company's proprietary 2D and 3D design data. It enables engineers to generate 3D models compliant with parametric constraints via natural language instructions, greatly lowering the technical threshold for modeling while improving design consistency. The system features multimodal input capabilities, recognizing text, sketches, and CAD files to achieve the intelligent generation of complex structures.

In the field of performance verification, IAT utilizes the Modulus framework to construct geometric neural network models trained on extensive 2D graphics and 3D CAD/CAE data. The average deviation between its predictive results and simulation data is controlled within 3%-5%, demonstrating high engineering practicality. This system provides rapid optimization suggestions ranging from concept design to performance evaluation, significantly accelerating the iteration speed of solutions.

Regarding large models, IAT has independently developed an automotive design knowledge large model, building a suite of vertical intelligent agent tools upon it. Through natural language dialogue, engineers can perform knowledge retrieval, rule-based reasoning, and process generation, gaining quick access to design specifications, parameter constraints, and historical use cases. 

Relying on this knowledge foundation, IAT has developed various applications—including design R&D assistants, review agents, solution generation agents, and regulation interpretation assistants, which effectively improves knowledge reuse rates and organizational collaboration efficiency.

Through this conference, IAT also seized the opportunity to exchange ideas with cross-domain experts from mechatronics, intelligent manufacturing, embodied intelligence, and other sectors. 

Looking ahead, IAT will continue to strengthen AI technology development and industrial collaboration, promoting artificial intelligence as a core engine for innovation within both the enterprise and the wider industry. Committed to "Efforts in Science for Better Cars," IAT will persistently inject technological momentum into China's journey from a manufacturing giant to a manufacturing powerhouse.

News