By 2025, the penetration rate of new energy vehicles in China will exceed 50% for the first time. Entering 2026, this data continues to rise and industry competition enters a white hot stage. From "3-year facelift" to "semi annual facelift", the iteration speed of the entire vehicle has doubled forward, but a core bottleneck has always been stuck in the upstream of the industry chain - the design and development of automotive molds, which still highly rely on the experience of engineers. The traditional design cycle of a set of molds is as long as 10 to 15 days, and engineers need to click and operate CAD design software up to 400000 times.

In Foshan, a national level specialized and innovative "little giant" enterprise is using artificial intelligence to break this dilemma. It compresses the mold design cycle from 10 days to 3 to 4 hours, increasing efficiency by more than ten times. This world's first stamping die AI intelligent design system comes from Dajie Intelligent Technology (Guangdong) Co., Ltd. In 2025, relevant technological achievements will be recognized as international advanced level by authoritative departments, and even internationally leading level in mold intelligent design technology, which will attract high attention at the NUMISHEET International Conference in Munich, Germany.

From 'tool software' to 'AI designer' 

 The story of Dajie Intelligence begins with a profound insight into the pain points of the industry.

Mold is known as the 'mother of industry', but its design heavily relies on traditional engineering experience, "said Yi Jianye, Vice General Manager of Dajie Intelligent. In the traditional mode, engineers mostly use commonly used foreign CAD software to complete 3D modeling, spending about 30% of their time making design decisions based on experience and 70% of their time on tedious mouse operations. The quality of design depends entirely on the personal experience and state of the engineer, and design errors are difficult to prevent

The greater crisis lies in the fact that industrial software is at the mercy of others. Industrial software for research and design has long been monopolized by foreign giants, with a localization rate of less than 10%. But the Dajie Intelligent team realized that artificial intelligence is bringing an opportunity for "overtaking on a different track" - not to catch up with decades of traditional software accumulation abroad, but to use AI to enable industrial software to complete a step upgrade.

In 2019, under the dual attraction of talent policies in Foshan and the high-end research platform of Jihua Laboratory, Professor Xie Hui, a doctoral supervisor at Hunan University, led a team to settle in Foshan, determined to tackle this tough nut.

The team focuses on the special characteristics of industrial data - private and closed source, technological timeliness, diversity and heterogeneity, and strong professionalism - and does not take the path of general large-scale models. Instead, they independently build the entire chain, from AI algorithms to cleaning massive engineering data, extracting professional knowledge to build a professional database, and then training vertical AI models to develop commercial AI systems.

A high-end algorithm development team of over 40 people and an engineering and technical team of over 50 people worked closely together, and after thousands of test cases and tens of thousands of problem validations, finally achieved a breakthrough.

In 2023, Dajie Intelligent launched commercial cooperation with local mold enterprises in Foshan, starting from a single segmented pain point scenario and overcoming technical difficulties in just over three months. Originally, they were also observing and questioning, but the efficiency increased by more than ten times, and the cooperation continued until the fifth phase, "Yi Jianye recalled.

The more the automotive industry becomes "saturated", the more "fragrant" AI becomes

The core products of Dajie Intelligent are the intelligent design system for stamping process and the intelligent design system for stamping die structure. After inputting part information, the system automatically recognizes features and completes the stamping process plan and process file output within a few minutes. We have created a process driven AI mold structure design, which originally required 10 to 15 days for mold structure design, but now can be completed in just a few hours.

This system is equivalent to a digital engineer with 3 to 5 years of experience, "said Yi Jianye. College students can operate the software for mold design after 1 to 2 months of training, and senior engineers only need to be responsible for review and optimization. The human resource allocation structure is deeply optimized.

More importantly, AI design brings not only time reduction. Traditional methods, in order to meet delivery deadlines, carry out mold structure design work in parallel when the process plan is uncertain, resulting in a large number of ineffective costs caused by modifications. "Yi Jianye calculated that" AI design can directly reduce cost waste by at least 30% to 40%. ”At the same time, industry and enterprise design standards are embedded in AI models, avoiding human errors and quality fluctuations caused by engineer turnover.

The current automotive industry is extremely "internal competition", and companies need to quickly launch new products. The more competitive the industry is, the more advanced technology iteration is needed, "said Yi Jianye frankly. Dajie Intelligent has established partnerships with leading car companies such as Dongfeng Lantu, BYD, FAW Volkswagen, Guangzhou Automobile Group, and Xiaopeng, transforming customers from "observing and questioning" to "actively seeking help".

At present, its cooperation model has been upgraded to tripartite joint innovation - open scenarios and standards for car companies, Dajie intelligent adaptation system, driving dozens of mold suppliers in the supply chain system to jointly use it, forming a highly cohesive intelligent design ecosystem.

Technology platformization, expanding to fields such as home appliances

Starting from automotive stamping molds, Dajie Intelligent's platform based intelligent design engine is extending to a wider range of industrial fields, including home appliances, aerospace industries, covering welding, die-casting, injection molding and other process areas.

Our AI design system has been platformized. ”Yi Jianye introduced that the platform is divided into three layers of architecture: the foundation layer, AI function layer, and application layer, encapsulating the common and reusable parts. Facing the new field of industrial equipment, it is only necessary to import industry knowledge, big data, and standards to develop exclusive intelligent design systems. At present, Dajie Intelligent has cooperated with Hisense in the field of home appliances and extended to the intelligent design of component products.