China develops AI robot capable of human farming tasks

A slender bionic arm weaves through tomato plants, with its 3D vision scanning the environment. Pausing at a flower, sensors on its fingertips assess pollen viscosity as algorithms process data in milliseconds.

Once adequate conditions are confirmed, high-frequency vibrations atomize the pollen, depositing it precisely on neighboring blossoms and completing artificial pollination with remarkable accuracy. The robot then navigates its way to the next plant autonomously, avoiding obstacles with a human-like level of awareness.

The intelligent agricultural robot was developed by researchers at Fudan University. Its system integrates multiple advanced technologies covering areas such as 3D vision perception, autonomous navigation, cloud-based decision-making and AI deep learning, culminating in a prime example of embodied intelligence.

Its seamless operations are the result of four years of development and four generations of prototypes — the work of a research team led by Shang Huiliang, an associate professor at the university.

The multi-functional robot handles the entire tomato cultivation process, including pollination, leaf pruning, fruit thinning and harvesting. Unlike previous single-function devices, this "thinking" machine can simulate human perception, decision-making and task execution.

The team's research journey began in 2021, when it identified agricultural robotics as a key development area. "We recognized that precision automation in agriculture presented both tremendous challenges and opportunities," Shang said.

The technical hurdles were daunting. Tasks like harvesting obscured fruits or navigating dense foliage, though simple for humans, proved extraordinarily difficult for machines.

Facing such complex challenges with limited personnel, Shang's team adopted a systematic approach to research and development. Early on in the project, the team leveraged the university's academic strengths, collaborating with optics and algorithm experts to solve fruit occlusion issues, while materials scientists and mechanical engineers developed flexible bionic arms.

Gradually, the team built up a compact, multidisciplinary research unit spanning the fields of mechanical engineering, electronics, automatic control, software development and AI.

And after four years of development, it evolved industrial robotic arms into the fourth-generation autonomous machines. "Our current model coordinates 'eyes, a brain, hands and feet' with embodied intelligence. It maintains over 90 percent pollination success even in challenging natural conditions," said Li Ruijiao, a researcher with the team.

Currently undergoing field tests at a Bright Food Group farm in Chongming, the model demonstrates remarkable efficiency. One robot can replace six laborers. This combination of performance and affordability has generated strong industry interest.

The team follows a demand-driven innovation model, with members having conducted extensive field research across 20 farms in Shanghai and provinces such as Qinghai, Guangdong and Hainan.

"We've inverted the traditional research and development process by starting from farmers' actual pain points. Developing hard technology requires constant problem-solving in real-world conditions," Li noted.