Zhou Juejia, Director of the Technology Committee Office at Xiaomi Corp. (left), introduces the self-developed "Training Kit" to a reporter. Zhou Juejia vividly recalls being stranded in a Canadian snowstorm years ago with a teacher from Shenzhen Institute of Information Technology. With nothing else to do, they discussed industry-education integration. "That teacher told me, 'Our courses haven't been updated for years. When we get back, I'll show you. Could some of your advanced technologies be introduced into the school?'" Since then, Zhou Juejia, who also serves as General Manager of Xiaomi Corp.'s Industrial Standard Research Department, has been contemplating how to deeply advance industry-education integration. He later visited the school's lab, where he found students still learning about bus protocols long obsolete in the industry. "What we need isn't taught in textbooks at all," Zhou remarked to the reporter, reflecting on the experience.
In early 2024, guided by the "Special Action for Industry-Education Integration in Vocational Education" initiated by the Ministry of Education's Vocational Education Development Center, companies like Xiaomi Corp., ZTE Corporation, Alibaba Cloud, and Volcano Engine have leveraged their respective industrial ecosystems to pave the way for talent cultivation in vocational colleges using real industrial demands and cutting-edge technologies. Outdated courses, technological disconnects, and impractical training have historically been the most tangible gaps between industry and education. Zhou Juejia noted that Xiaomi Corp.'s business now spans numerous fields, including smartphones, AIoT (Artificial Intelligence of Things), the human-vehicle-home ecosystem, smart manufacturing, and automobiles. "However, a vast amount of the technologies, systems, and development processes widely used within our company are completely absent from traditional textbooks," he said.
Under the "Special Action for Industry-Education Integration in Vocational Education" framework, 147 institutions were selected for Xiaomi Corp.'s new-generation intelligent technology industry-education integration promotion project in 2025. In just over half a year, Xiaomi Corp. and its ecosystem partners have absorbed over 4,000 graduates from cooperating schools. How can technology be transformed into curricula and training into practical experience to bridge the blockage between institutional education and industrial needs? Xiaomi Corp.'s answer is a self-developed "Training Kit." Zhou Juejia explained that the kit's初衷 was to create a tool that makes teaching easier by incorporating all of Xiaomi Corp.'s technologies. In his words, this Lego-like box aims to "shorten the distance between the industrial frontier and the vocational education classroom." Through the kit, Xiaomi Corp. translates its abstract technology ecosystem into a tangible, hands-on operational process delivered to institutions.
During the interview, Zhou Juejia displayed the finished "Training Kit," officially named the "Xiaomi Corp. AIoT Development Platform." Roughly the size of a standard carry-on suitcase, it contains several rectangular functional modules integrating all functionalities of Xiaomi Corp.'s whole-house smart system. These modules can be combined via magnetic grooves and connected to the terminal built into the kit. "With reliable hardware, students can focus more on algorithms and logic," Zhou stated. Students can combine modules for computing, communication, sensing, and execution in an experience akin to "building with Legos." Furthermore, these "semi-finished products in appearance" can directly connect to the "Mi Home App" at the internal hardware level, fully interoperable with market terminal devices, allowing students to conduct development and debugging in real-world scenarios.
Pan Chen, Dean of the School of Information Engineering at Jinan Engineering Vocational and Technical College, told the reporter that the project's core is establishing a genuine "Xiaomi Corp. Innovation Practice Base." The kit provided by Xiaomi Corp. miniaturizes and pedagogically adapts complex AIoT systems. Students learn the design, installation,调试, and maintenance of smart homes here, with all operational logic完全一致 with Xiaomi Corp.'s frontline stores. "In the past, our students weren't 'plug-and-play' when they joined companies; there was a technological gap. Now, training with this kit significantly shortens their adaptation time after taking up their posts, potentially reducing it from a month to a week," Pan said.
Hebei Software Vocational and Technical College is also integrating Xiaomi Corp.'s "human-vehicle-home ecosystem" into its classrooms. The school currently has 22 AIoT training kits. Gao Xiuyan, Director of the Software Engineering Department, noted that while students work on smart home networking with the kits, they can combine corresponding module components and even simulate signal屏蔽 and noise caused by walls to study signal attenuation.
Meanwhile, against the backdrop of AI rapidly empowering various industries, some companies are choosing to跨越 the spatiotemporal limitations of industry-education integration by developing specialized vertical models tailored for vocational education. Since 2025, ZTE has collaborated with 53 institutions on corpus collection. The corporate side provides cutting-edge technical standards, desensitized industry data, and algorithm engineering capabilities, while the academic side aggregates high-quality teaching resources, teaching behavior data, and typical application scenarios. Once trained and deployed into the institutions' teaching and research environments, the models can support the development of AI courses, innovation in intelligent teaching tools, and applications in teaching management scenarios.
Similarly, ByteDance, leveraging the technology and computing power of "Volcano Engine," cooperates with institutions on AI-empowered vocational education. Multiple AI application products are used in intelligent teaching processes, possessing capabilities such as training and inference for educational vertical models, intelligent interaction in education and teaching, and intelligent search and recommendation of teaching resources. In the future, backed by ByteDance's powerful large model iteration and升级 capabilities, more educational AI scenarios will be continuously explored until implementation. Alibaba Cloud, relying on its multimodal data automated annotation platform, assists institutions in organizing datasets that meet model training standards and utilizes vertical industry models to develop specialized professional courses.
What is a "vertical model"? Lu Xiaofei, General Manager of the Education Industry at ZTE, explained that the AI dialogue platforms people use daily are mostly general models capable of answering broad questions but may produce "hallucinations" in professional fields. Vertical models are specifically trained on语料 from particular domains, enhancing the model's professional capability and accuracy. From the perspective of vocational college cooperation, companies leverage their technological advantages to adapt and融合 internal industrial语料 with academic teaching语料, creating vertical models that can more precisely fit actual teaching, smoothing out the information gap between industry and education.
In fact, for the internet and communications industry, AI vertical models have become an indispensable core "brain" at the industrial frontier. Lu Xiaofei told the reporter that the vertical models deployed within ZTE now cover大部分 areas like code writing, network diagnosis, and smart manufacturing. The production line automation rate at ZTE's Nanjing Smart Binjiang 5G Factory has exceeded 85%. "We used to use 5G to manufacture 5G; now we use AI to manufacture AI," Lu said. In his view, the accelerated transformation and upgrading of industry is also forcing a transformation in vocational education talent cultivation. He observes that new technologies may iterate every few months, and the emergence of AI can also help education achieve personalized, tailored teaching that follows industry trends, encouraging students to enhance their practical abilities in different fields.
Although the tracks and models of different companies vary, the underlying logic of industry-education integration is highly consistent: driven by industrial demand, transforming a company's most core, authentic, and cutting-edge technologies into educational resources that are teachable, trainable, and certifiable within vocational institutions. As Zhou Juejia mentioned, cultivating "specialized yet versatile" talent is a key focus for vocational education today, especially in the rapidly changing AI era where "mastering learning methods is more important than mastering knowledge itself."
Comments