[New Leap in Productivity·Power of Source] “AI Robot Laboratory” develops new drugs in Zhangjiang Robot Valley
Have you ever seen a scene where robots develop drugs? Recently, reporters from Liberation Daily and Shangguan News visited the AI Robot Laboratory located in Zhangjiang Robot Valley. More than 100 robots are operating in the laboratory to carry out high-throughput drug research and development experiments. "We want to use robotics and artificial intelligence technology to empower the research and development of new drugs and form new productivity in the pharmaceutical industry." said Wang Mingtai, vice president of Jingtai Technology.
Looking at the entire Zhangjiang Robot Valley, it is located in the core area of Zhangjiang Science City. With humanoid robots as the core, medical robots, industrial robots and service robots as the focus, and breakthroughs in key robot components and key control software, it is making every effort to promote the robot industry. "Chain development", create a world-class robot industry cluster, and build the world's leading robot industry and innovation ecosystem.
Flexible robotic arms conduct experiments and collect data to train large models
Jingtai Technology is a technology company based on quantum physics, artificial intelligence empowerment and robot-driven technology. It helps biomedicine, new materials and other industries improve quality and efficiency, and cultivate and develop new productivity. The company has established industry-leading innovation R&D centers in Shanghai, Shenzhen, and Beijing, and built a large-scale AI robot laboratory in Shanghai.
Walking into this unmanned laboratory, the reporter saw robot workstations one after another. Through the glass of the workstation, one could see robotic arms manipulating test tubes one by one and conducting chemical experiments in an orderly manner. "This laboratory has a lot of know-how. The entire operating system and scheduling system are independently developed by us. Hardware equipment such as robotic arms are jointly developed with partners." Wang Mingtai said that these robot workstations have high flexibility and can meet the needs of It meets the experimental needs of different research and development scenarios, and keeps gentle and dexterous movements when doing experiments. Not only will it not break the test tube, it can also complete delicate operations such as screwing bottle caps, pipetting, shaking reactions, and filtration detection.
Each workstation is also equipped with vision, weighing and other modules to collect data generated in robot experiments in real time. These data are the raw materials for training large vertical AI models. With the rise of general-purpose large models such as ChatGPT, many companies and scientific research institutions have begun to develop vertical large models suitable for various industries. These models require data for training. Jingtai Technology is also exploring this cutting-edge field. With the expansion of AI robot laboratory application scenarios, it has created first-class conditions for data collection in the biomedical industry, which is expected to enable large models to predict experimental results and discover new drugs.
According to reports, Jingtai Technology has developed the ProteinGPT model, which has made breakthroughs in solving the "one-click drug" problem for macromolecules. This model is covering scientific research scenarios such as antibody discovery, antibody transformation, antigen design, and protein de novo design. ProteinGPT not only contributes to the development of macromolecule biopharmaceuticals, but also has broad application prospects in other industrial fields related to protein macromolecules.
Cooperate with domestic and foreign pharmaceutical companies to shorten the development cycle of new drugs
In recent years, "artificial intelligence biomedicine" has become a popular track in the international technology and industrial fields. As the leading company in this track, Jingtai Technology has reached cooperation with many multinational pharmaceutical companies, well-known domestic pharmaceutical companies, and pioneer Biotech companies, such as Pfizer, Eli Lilly, Singapore National Drug Research and Development Platform EDDC, Yangtze River Life Sciences, and Chia Tai Tianqing wait. Among the top 20 multinational pharmaceutical companies in the world, 16 have established cooperative relationships with this Chinese company.
During the development process of PAXLOVID, the world's first anti-COVID-19 small molecule oral drug approved by the FDA, Jingtai Technology teamed up with Pfizer and used its independently developed AI prediction algorithm combined with experimental verification to significantly shorten the new drug development cycle, taking only 6 weeks. It took time to confirm the dominant crystalline form of the drug.
Computational biology experiments using AI prediction algorithms are called "dry experiments", while traditional biological experiments are called "wet experiments". The combination of "dry experiment" prediction and "wet experiment" verification can help scientific researchers achieve higher-quality drug research results and significantly improve scientific research efficiency. For example, in the new drug discovery process, AI algorithms can quickly screen out potentially active molecules from hundreds of thousands of compounds, providing source molecules for "wet experiment" drug discovery, which is of great significance to the development of innovative drugs. In May last year, Jingtai and Eli Lilly signed a cooperation in the discovery of new AI small molecule drugs. The total revenue from advance payments and milestones reached US$250 million, setting a new record for the amount of a single AI pharmaceutical pipeline in China.
The first floor of Jingtai Technology's Shanghai headquarters is the AI robot laboratory, and the second floor is the traditional "wet laboratory." The advantages of robots in conducting experiments are high continuity and large throughput. They can work 24 hours a day, 7 days a week, and can carry out dozens or even hundreds of experiments at the same time. They are suitable for common chemical experiments with simple steps. On the contrary, the throughput of experiments conducted by R&D personnel is relatively small, which is suitable for carrying out experiments with multiple reaction steps, complex reaction types, and higher technical and experience requirements, including compound synthesis, separation and purification, reaction monitoring, and compound structure confirmation.
The effective connection between the two laboratories and the close cooperation between chemical robots and scientific researchers have improved the efficiency of drug research and development, and will also provide more and more high-quality data for "dry experiments", making the AI "brain" smarter and smarter.
![](https://a5qu.com/s/user/default.webp)