Shenzhen issues the "Artificial Intelligence+ Advanced Manufacturing Action Plan (2026–2027)"

Notice on the Issuance of the “Shenzhen City ‘Artificial Intelligence+’ Advanced Manufacturing Action Plan (2026-2027)”

To all relevant units:

In order to thoroughly implement the relevant documents of the State Council on the deep implementation of the “Artificial Intelligence+” initiative, our bureau has formulated the “Shenzhen City ‘Artificial Intelligence+’ Advanced Manufacturing Action Plan (2026-2027)”. It is now issued for implementation. Please organize and carry out its implementation conscientiously.

Shenzhen Municipal Bureau of Industry and Information Technology

February 9, 2026

Shenzhen City ‘Artificial Intelligence+’ Advanced Manufacturing Action Plan (2026-2027)

To thoroughly study and implement the spirit of the 20th National Congress of the Communist Party of China and the successive plenary sessions, and to earnestly implement the “Opinions of the State Council on Deepening the Implementation of the ‘Artificial Intelligence+’ Action,” seize the historical opportunity at the intersection of intelligence and industrialization, accelerate the deep integration of artificial intelligence technology with the entire process and all elements of manufacturing, fully serve and support new industrialization, this plan is formulated.

1. Overall Requirements

Anchored in the strategic goal of achieving new industrialization, promote the acceleration of AI empowering manufacturing, and deepen its penetration into research and development design, production management, manufacturing operations, operation management, and supply chain management. Promote the intelligent development of all manufacturing elements, achieve comprehensive, deep, and high-level AI empowerment of new industrialization. By 2027, in the field of “Artificial Intelligence+” advanced manufacturing, establish a national AI application pilot base (consumer mobile terminal direction), build an industrial intelligent agent innovation center, form an industrial knowledge alliance, open up 100 application scenarios, create 100 vertical industry models and industrial intelligent agents, promote 100 demonstration applications, and form a development pattern of “one base, one center, one alliance, hundreds of scenarios, multiple applications.” Drive the renewal and upgrading of traditional industries, the leapfrogging of emerging industries, and accelerate the advancement of new industrialization.

2. Build Key Supporting Platforms

(1) Build an Industrial Intelligent Agent Innovation Center. Accelerate the construction of provincial-level industrial intelligent agent innovation centers and strive for layout in national manufacturing innovation centers. Focus on industrial scenarios such as “digital workers,” support R&D of industrial intelligent agents with environmental perception, autonomous decision-making, and dynamic adaptation capabilities, concentrate on scenarios like R&D design, manufacturing, and supply chain management, gather high-level intelligent agent application developers, establish a demand-supply matching platform for industrial intelligent agents, build a self-controlled technical foundation, develop dedicated toolchains for industrial intelligent agents, create an open and shared ecosystem, and improve collaboration levels in complex industrial scenarios.

(2) Develop Industrial Software and Industrial Knowledge Alliance. Support enterprises in transforming industrial knowledge and industry experience into standardized models, focus on key industrial software such as operating systems, CAD, CAE, EDA, and their large model adaptation and development, support industrial large model industrialization in key scenarios, and form industry-leading autonomous industrial software products. Recognize the trend of miniaturization of industrial large models, support the use of pruning, quantization, distillation, and other model compression techniques to develop lightweight scenario-specific industrial small models, enabling low-latency edge decision-making and inclusive deployment. Build a platform for co-constructing industrial knowledge, gather efforts from enterprises, universities, and research institutions, construct industry-level knowledge covering R&D design, manufacturing, and supply chain management, sediment core knowledge entities and relationships, and form a large-scale industrial knowledge database. Establish an open community platform, encourage leading enterprises to open application scenarios, lower the threshold for SME intelligence, provide industrial knowledge sharing, AI application development toolkits, and other inclusive services, and foster an ecosystem of integrated development among large, medium, and small enterprises.

3. Empower Key Industry Clusters

(3) AI Empowerment of Electronic Information Manufacturing. Build public service capabilities for open-source algorithms, data sharing, and computing power collaboration, integrate technical resources and industry data, reduce the barriers for SMEs to upgrade intelligently, and enhance the penetration and application depth of industry intelligence. Strengthen the leading role of top enterprises, jointly explore potential application scenarios with upstream and downstream industry chain enterprises, support AI in core links such as product design, testing, operation management, quality inspection, safety production, and data analysis, and create benchmark demonstration projects. Focus on terminal product innovation and upgrading, support R&D and innovation of AI smartphones, AI glasses, AI+ trendy gadgets, AI+ smart screens, etc., and drive technological iteration through product innovation to cultivate new growth points.

(4) AI Empowerment of Semiconductors and Integrated Circuits. Promote AI technology application in key links of the semiconductor industry chain, optimize efficiency in chip design, software coding, and other areas using AI. Strengthen the semiconductor industry with AI chips as a breakthrough, develop high-performance, high-efficiency dedicated SoC main control chips for AI smartphones, AI glasses, and intelligent robots, support new architectures such as integrated storage and computing, and in-processor computing. For the trillion-yuan new energy vehicle market, support domestically produced high-end AI chips for automotive-grade 14nm and below, intelligent driving, smart cockpit SoCs, domain controllers MCU, and central domain control SoCs/MPUs as replacements.

(5) AI Empowerment of Automotive Manufacturing. Pilot applications of intelligent connected vehicles with “vehicle-road-cloud integration,” increase AI empowerment across the entire automotive industry chain. In collaborative design, intelligently manage component resources, recommend optimal component information, use AI algorithms to simulate and automatically match and clean material attributes, and improve high-precision mesh division to enhance R&D efficiency. In manufacturing, coordinate resources intelligently, optimize manufacturing resource allocation, and promote efficient utilization of idle manufacturing resources. In inspection and testing, use intelligent scheduling for equipment task distribution, data analysis, automatic problem detection, and intelligent data feedback to improve product yield. In packaging and verification, intelligently identify and match demand data, flow data, and resource data, manage warehouse logistics, and optimize material batching and collaborative configuration.

(6) AI Empowerment of Robots. Support the R&D of multimodal interaction technologies such as world models, vision-haptic-language-action (VTLA), build embodied intelligent base large models with interaction, prediction, and decision-making functions, and develop training and inference systems. Cultivate long-sequence reasoning and autonomous learning capabilities to support efficient handling of cross-scenario tasks. Strengthen scenario resource coordination, support the construction of embodied intelligent technology test fields, open up segmented scenarios such as welding, assembly, spraying, and handling in industrial manufacturing, and implement applications. Improve intelligent operations in hazardous and harsh environments, promote robots into factories, workshops, warehouses, ports, and parks.

(7) AI Empowerment of High-Performance Materials. Support AI-enabled manufacturing of high-performance materials, encourage the use of AI for dynamic process parameter optimization and precise control, and promote overall intelligent production processes. Build an AI high-performance materials supply and demand platform, create a demand-driven, rapidly iterative, resilient, and efficient materials industry ecosystem, and develop flexible manufacturing, agile response, and service innovation models. In design and screening, mobilize relevant units to participate in high-performance materials data center construction, use machine learning algorithms to predict the structural properties of polymers, metals, and inorganic non-metals, and assist R&D personnel in selecting high-performance materials. In process and pathway optimization, train large models with domain knowledge bases for chemical synthesis to find optimal synthesis routes. In performance prediction, use AI models to forecast properties such as elasticity and thermal conductivity. In experimental guidance, leverage code-generation large models combined with simulation platforms and intelligent robots to automate simulation experiments.

(8) AI Empowerment of Low-Altitude Economy. Establish an autonomous capability evolution system for drones, build intelligent simulation platforms, create low-altitude digital twin systems, deeply integrate AI technologies to support perception and decision-making simulation and testing, strengthen autonomous task execution, and cultivate aerial embodied intelligence. Construct an “Aerial Smart Road System” to support intelligent airspace design, route planning, achieve full-airspace perception, drone management, and multi-drone autonomous coordination. Enable applications such as park, river, reservoir, shoreline inspections, manned flight, logistics, low-altitude tourism, aerial sports, flight training, power line patrols, port inspections, aerial mapping, agricultural and forestry pest control, etc., improving low-altitude resource scheduling efficiency and coordinated operation.

(9) AI Empowerment of Medicine and Medical Devices. Accelerate innovation and transformation in drug development, cell and gene therapy, and precision medicine services. Promote AI technology in core areas such as new target discovery and validation, drug design, ultra-high-throughput screening, DNA-encoded compound libraries, computer-aided drug design and virtual screening, and gene site screening related to drug therapy. Support the construction of major AI drug R&D platforms, strengthen resource integration, and accelerate the deep integration of AI and biotechnology (AI+BT). Strengthen the leading role of large model enterprises and high-end medical device companies, collaborate across the industry chain for joint innovation in medical equipment and key components, open up large-scale real application scenarios such as medical imaging-assisted diagnosis, and promote high-end, intelligent upgrades of medical devices, creating benchmark “AI+Medical Devices” applications.

(10) AI Empowerment of Traditional Advantage Industries. Explore new paths for the optimization and upgrading of traditional industries, encourage deep application of AI technologies such as large models, intelligent agents, machine learning, computer vision, and natural language processing in industries like apparel, watches, glasses, jewelry, furniture, and leather. Focus on generative AI design empowerment, small-batch flexible production, C2M reverse customization, and supply chain intelligent scheduling. Develop a batch of vertical large models and intelligent agent upgrade benchmarks, promote transformation from scale-driven to “creativity + efficiency + personalization,” improve industry productivity and product quality, reduce production costs, and foster high-end, intelligent, green, integrated, and international development of traditional industries, cultivating new growth poles for high-quality industrial development.

4. Strengthen Work Guarantee

(11) Policy Support and Element Support. Increase financial support for “Artificial Intelligence+” advanced manufacturing, encourage enterprises to actively participate in “unveiling and appointing,” promote policies and technological reforms for AI empowering new industrialization, industrial internet, digital transformation, and intelligent manufacturing, accelerate the industrialization of innovative achievements, further strengthen collaborative efforts, and promote the digital, networked, and intelligent transformation of manufacturing.

(12) Expand Scenario Opening and Demand-Supply Docking. Deepen the scenario demand-supply matching mechanism for “Artificial Intelligence+” advanced manufacturing, establish application scenario open centers at municipal and district levels, hold series of demand-supply docking activities, release scenario demand lists, support leading enterprises in fully opening product design, intelligent testing, large-scale customization, and intelligent logistics scenarios, explore and develop a batch of high-potential, high-benefit, high-value new scenarios, and give full play to social organizations as bridges and coordinators to support solution providers and key industry enterprises in overcoming AI application challenges.

(13) Conduct Industry Training and Demonstration Promotion. Regularly hold “Artificial Intelligence+” advanced manufacturing industry training focusing on AI frontier technologies, collect key AI solutions addressing industry pain points, form typical cases, promote AI demonstration applications and excellent solutions, strengthen the demonstration and leading role of industry benchmarks, and create a strong atmosphere for AI empowering new industrialization.

(Source: Shenzhen Municipal Bureau of Industry and Information Technology)