When it comes to synthesizing and developing critical materials for the new energy and semiconductor industries, humans may soon step aside. At Tan Kah Kee Innovation Laboratory, experiments run continuously and efficiently in a 200-square-meter laboratory with no one inside—at several times the efficiency of manual experiments.
A team at Tan Kah Kee Innovation Laboratory has built a Smart Laboratory for critical materials serving strategic emerging industries such as new energy, new materials, and semiconductors, providing AI for Science solutions across multiple fields. From material synthesis and testing to analysis and refinement, the process no longer requires human intervention, forming a fully automated and intelligent closed loop. On this basis, Tan Kah Kee Innovation Laboratory incubated EVA AI for Science (hereinafter referred to as “EVA AI for Science”) to bring the technology to industrial applications. The reporter recently visited the EVA AI for Science laboratory.
Breaking Free from Human Intervention
This laboratory was designed specifically for robots: 6 experiments run simultaneously in the synthesis system; passageways and workbenches are arranged for robotic operation; material-handling robots move freely through the space, while robotic arms operate equipment methodically…
Recently, reporters at the EVA AI for Science laboratory witnessed an unprecedented scene: it is currently China's first Autonomous Laboratory dedicated to electrochemical energy materials R&D. Unlike the bustling people-and-equipment setup of a conventional laboratory, this facility normally has no people on site and requires only one engineer to occasionally handle control and analysis from the central control console.
“What you see is simply an unmanned operation, but behind it, the robots do far more than replace manual laboratory work. They create an R&D closed loop of high-throughput synthesis, automated characterization, and intelligent decision-making.” Zhang Yunpeng, general manager of EVA AI for Science, told the reporter. After the laboratory synthesizes a material, it must undergo a “physical examination” like a person; then, based on the performance metrics in the “report,” intelligent analysis determines how to synthesize the material better next time.
The central control console nearby serves as the brain of the entire laboratory. All completed experiment data is aggregated here, then analyzed by AI models to determine what experiment should be run next.
Such Smart Laboratories eliminate dependence on people during experimentation and decision-making. With advanced AI, robots gain human-like observational ability and dexterity; even without guidance, they can independently determine and perform tests. This not only increases speed but also reduces errors that may arise from manual operation.
“This is a qualitative leap,” Zhang Yunpeng said. “This AI-powered, high-throughput, automated closed-loop Intelligent Materials R&D System not only turns materials development into an ‘autonomous intelligent R&D’ model similar to self-driving, but also greatly improves efficiency.”
The Smart Laboratory has also developed an AI for Science R&D System for lithium-ion batteries, comprising two modules: automated lithium-ion battery assembly and high-throughput testing. This is an exploration by EVA AI for Science in collaboration with a leading new-energy company, following the model of “industry defines the problems, and the laboratory solves them.”
“Traditional lithium-ion battery materials R&D faces challenges such as high costs, low energy density, and short cycle life. Bringing in AI will deliver a revolutionary leap in battery materials R&D,” Zhang Yunpeng explained. “The laboratory’s Intelligent Energy Materials R&D System covers the full R&D chain, from automated lithium-ion battery assembly to high-throughput testing, enabling intelligent fabrication of lithium-ion batteries.”
Compared with manual assembly of lithium-ion batteries, it provides more precise control over the proportions of different materials and achieves assembly efficiency 5 times that of manual processes.
Core Equipment Is Fully Domestically Produced
Founded in March this year, the young company secured tens of millions of yuan in angel financing from a professional investment institution just last month.
Why is this Smart Laboratory so exceptional? One key challenge is that people who understand chemistry may not necessarily understand automation and AI, while those who understand automation may not know AI or chemistry well. But the project’s leader, Professor Hong Wenjing, deputy director of Tan Kah Kee Innovation Laboratory, brought chemistry, automation, AI, and other disciplines together. Starting from the ground up, he recruited engineers in mechanical engineering, software, electrical controls, and other fields to build a cross-disciplinary R&D team.
“First, we need to define what materials to develop and what R&D equipment is required; then electrical-control, mechanical, and software algorithm engineers turn the concept into a deployable solution,” Zhang Yunpeng explained. “From forming the team and building the laboratory to closing the loop for specific material systems, the research team spent more than two years.”
More importantly, the laboratory requires all core instruments and equipment to be domestically produced, thereby creating an ecosystem of Chinese scientific instruments. On the one hand, drawing on Xiamen University’s research foundation in scientific instruments, the R&D team has developed a series of intelligent scientific instruments and equipment in-house. On the other hand, it is collaborating with multiple domestic manufacturers to co-develop NMR, Raman, and other spectroscopic instruments, coordinating with them to provide open interfaces for the laboratory.
“We must solve real problems for China’s industrial technological innovation—and solve them for real,” Hong Wenjing said. As an interdisciplinary researcher in chemistry, chemical engineering, and artificial intelligence, he leads an engineering R&D team of more than 20 people. Starting from the ground up, the team connects the chemistry, chemical engineering, materials, artificial intelligence, and scientific-instrument value chains to create an ecosystem for intelligent R&D and innovation.
Producing research achievements like these requires interdisciplinary integration as a prerequisite. “Within a 1-kilometer radius of Tan Kah Kee Innovation Laboratory, one can find researchers from more than a dozen different disciplines,” Hong Wenjing said. “Interdisciplinary collaboration has made research more engaging and broadened the horizons of researchers.”
More importantly, EVA AI for Science aims to free up more scientists’ time through AI, advance AI for Science R&D in materials, and foster next-generation productive capabilities. “We are building a general-purpose AI laboratory. Like assembling LEGO bricks, our modular AI laboratory can be configured to meet different needs,” Zhang Yunpeng said. As a provider of Intelligent R&D Systems and Smart Laboratory solutions, EVA AI for Science will deliver AI for Science solutions for critical materials in strategic industries—including new energy, new materials, and semiconductors—tailored to customers’ R&D needs.
