Chinese Researchers Use AI to Transform Hydroelectric Dam Project into World’s Largest 3D Printer
The Yangku Dam, standing at 180 meters tall, is set to be constructed layer by layer using excavators, trucks, bulldozers, concrete mixers, and autonomous vehicles. All machinery will be controlled by artificial intelligence (AI), mirroring the additive manufacturing processes utilized in 3D printing. Once completed in 2024, the Yangku Dam will generate nearly 5 billion kilowatt-hours of electricity annually from the upper reaches of the Yellow River, flowing into Henan, home to 100 million people. The electricity will be transmitted through a 1,500 km long high-voltage power line.
The Yangku Dam will be ready for operation in two years. (Photo: Weibo).
According to Liu Tianyun, the lead scientist on the project, in a study published in Tsinghua University’s journal, the construction process of the dam and 3D printing “are fundamentally identical.” After years of development and testing, the technology for large-scale infrastructure 3D printing has matured enough for widespread application and “liberates humans from heavy and dangerous manual labor.”
Liu, a researcher at the Hydrology and Engineering Laboratory at Tsinghua University, and his team conceived the idea to “print” large-scale construction projects about a decade ago. They envisioned the entire construction site transforming into a giant printer, with various automated machines working seamlessly together like different parts of a machine.
The initial development of 3D printing was aimed at producing components from valuable materials with minimal waste. Compared to traditional cutting and milling methods, 3D printing generates less waste. Since then, some architects have begun applying this technology to buildings, although these projects have been relatively small in scale. The first 3D printed office building is the headquarters of the Dubai Future Foundation, standing just 6 meters tall.
Chinese civil engineers are no strangers to AI, which has been widely used to construct the Baihetan Dam, the second largest dam in the world, in just four years. To date, this technology has primarily played a coordinating role in projects. Previous experiments with technology in various construction projects have shown that intelligent machines can outperform humans “especially in harsh and dangerous environments,” the Liu research team stated.
According to media reports, construction of the Yangku Dam began late last year in the Hainan Tibetan Autonomous Prefecture in Qinghai Province. After dividing the dam’s computer model into multiple layers, AI will assign a fleet of robots to gradually build each layer. Autonomous excavators can identify and load materials from the site onto automated trucks, some of which are electric.
Following an optimal route calculated by AI, the trucks will deliver materials to the corresponding locations at the right time. Robotic bulldozers and concrete mixers will use these materials to construct each layer of the dam structure. Autonomous rollers will compact the newly poured layers, equipped with sensors. AI utilizes these sensors to monitor construction quality by analyzing ground vibrations and other data.
Breakthroughs in AI technology, including enhanced deep learning, mean that machines can now recognize nearly all objects on-site, respond to unexpected situations in a constantly changing environment, and perform a variety of flexible tasks. These machines also avoid common human errors. According to Liu, truck drivers often deliver materials to the wrong locations, while collisions and strong vibrations can hinder human operators from keeping rollers on a perfectly straight path. Furthermore, most workers cannot read technical blueprints. However, the greatest advantage of machines is their ability to work in life-threatening environments without suffering from headaches due to lack of oxygen or exhaustion after working continuously throughout the day.
Not all tasks in the dam construction process are handled by machines. The research team noted that rock excavation from nearby mountains is managed by workers due to the complexity of the task. They mentioned that this technology could also be applied in other infrastructure projects such as airport construction and road building.
An anonymous civil engineering scientist in Nanning stated that 3D printing technology has some limitations but holds great potential for future applications. “It cannot print a structure that includes different materials like reinforced concrete and cement. But a fleet of construction robots can compensate for the significant decline in labor due to slow birth rates,” they noted.
