Beijing has formally released an implementation plan to accelerate the application of artificial intelligence in scientific research, with a goal of establishing a globally influential innovation hub in this field by 2028.
The "Beijing Implementation Plan for Accelerating AI-Enabled Scientific Research (2026-2028)" outlines the city's strategy to achieve major original breakthroughs in the fundamental theories and key technologies of AI for science.
The plan aims to build a science and intelligence innovation hub with global reach, providing strong support for the development of the Beijing-Tianjin-Hebei International Science and Technology Innovation Center.
The strategy is structured around five key directions: autonomous laboratory systems, high-value application scenarios, scientific model systems, scientific data foundations, and the innovation ecosystem.
This framework encompasses 18 specific tasks, including the construction of autonomous labs, the promotion of intelligent upgrades for major scientific facilities and research instruments, and the development of scientific research agents and toolkits.
Establishing the Autonomous Laboratory Framework
One of the primary goals is to accelerate the development of an autonomous laboratory system. This involves coordinating the construction of self-operating labs, the intelligent upgrading of major scientific infrastructure and instruments, and the research and development of scientific agents and tools.
The objective is to create intelligent experimental platforms capable of full-process autonomous operation, establishing a new "dry-wet" closed-loop research model that integrates computation and experimentation.
The plan seeks to complete the iterative upgrade and open sharing of a batch of major scientific and technological infrastructures.
Promoting High-Value Benchmark Applications
The initiative will focus on driving benchmark applications in six key fields: high-energy physics, materials science, healthcare and medicine, life sciences, quantum technology, and biological breeding.
By deploying intelligent scientific research tasks in these areas, the plan aims to create a series of exemplary scenarios to accelerate the output of innovative results.
Applying AI to fundamental research and industrial R&D is expected to shorten development cycles and reduce experimental costs.
Strengthening the Layered Scientific Model System
The strategy involves a progressive, three-tiered layout: fundamental theory, general-purpose scientific foundation models, and specialized models for specific sub-fields.
It aims to break through core technologies such as multi-scale modeling and novel neural networks, and to deeply integrate AI models embedded with physical laws with traditional scientific computing.
A coordination mechanism between foundation models and specialized models will be established, striving to build a collaborative scientific model system of "foundation models + professional models" that meets world-class standards.
Building a Standardized Scientific Data Foundation
The plan calls for advancing the construction of scientific data centers in a coordinated manner and improving Beijing's scientific data management system to aggregate and integrate multi-source research data resources.
It also includes building an intelligent scientific data processing tool system to create high-quality, standardized, and reusable foundational discipline-specific scientific datasets.
Fostering an Open Innovation Ecosystem
The final pillar focuses on cultivating an open innovation ecosystem by coordinating four major measures: attracting and nurturing interdisciplinary talent, promoting international scientific exchange and cooperation, fostering regional collaboration, and ensuring computing power supply.
This includes setting up global academic exchange platforms, coordinating and dispatching Beijing's computing resources, and building an integrated support system for talent, computing power, and open cooperation.
The goal is to form a scientific intelligence innovation ecosystem characterized by the concentration of high-end talent, open-source sharing, and deep integration of industry, academia, research, and application.
An official from the Beijing Municipal Science and Technology Commission and the Administrative Committee of Zhongguancun stated that the plan leverages Beijing's integrated advantages in education, technology, talent, AI industry, and scientific research resources.
It constructs a full-chain promotion system encompassing "models, data, agents, autonomous labs, scenarios, and ecology," coordinating various policy tools and innovation resources at all levels with the goal of building a global hub for scientific intelligence innovation.
The next steps will involve taking the lead in implementing the policy, organizing collaborative research efforts among Beijing's leading research institutes, enterprises, and new R&D institutions, and promoting the itemized and project-based execution of all tasks to continuously improve the scientific intelligence innovation ecosystem.
Comments