China has deployed an artificial intelligence-driven 'fish facial recognition' system on a fishway at the Zangmu Hydropower Station on the Yarlung Tsangpo River in Tibet, enabling round-the-clock automated species tracking that replaces labour-intensive manual counts. The system identifies fish species unique to the Tibet Autonomous Region and monitors their passage continuously, according to a report published by Global Times on Tuesday, 10 August 2026.

How the AI system works

The AI-based identification platform operates along the fishway — commonly called a fish ladder — built into the 116-metre-tall (381 feet) Zangmu Hydropower Station, which has been operational since 2015 on the middle reaches of the Yarlung Tsangpo River. Rather than relying on periodic human observation, the system runs uninterrupted, logging every species that passes through the passage. The technology applies image-recognition algorithms trained on regionally specific aquatic fauna, effectively giving each fish a biometric profile.

Why it matters for conservation

Since the dam began operating, its fishway has helped approximately 570,000 rare fish swim upstream — a figure that underscores the ecological pressure large hydropower infrastructure places on migratory species. Yang Xiaocheng, Director of Safety and Environmental Protection at the dam, said the monitoring data points to a healthy ecosystem. 'From the monitoring results, the fish population in the middle reaches of the Yarlung Tsangpo is very stable and has not decreased because of the construction of the hydropower station,' Yang said. He added: 'On the contrary, the aquatic ecology and animal-plant life in the entire middle gorge section of the Yarlung Tsangpo River have shown positive changes.'

The competitive backdrop

Automated fish-monitoring technology is an emerging field globally. Institutions including the Alaska Fisheries Science Centre and private firms such as Innovasea in Canada and Nova Scotia Power have developed acoustic and video-based fish-passage monitoring systems. The US Department of Energy has also funded fishway research. China's deployment of AI facial-recognition methodology — a technique the country has refined extensively in urban surveillance — represents a novel crossover application into wildlife conservation at scale.

Hydropower's environmental trade-off

Hydropower facilities provide flood protection and a reliable renewable energy source that can offset intermittent generation from solar and wind installations. However, they restructure natural water bodies and fragment migratory corridors for aquatic species, posing significant environmental challenges. The Yarlung Tsangpo, which flows downstream into Bangladesh and India as the Brahmaputra, is ecologically sensitive and home to species such as Glyptosternon maculatum. Downstream nations have long raised concerns about the cumulative impact of dam construction in the upper basin.

What's next

The application of AI-based biometric identification to fishway monitoring could inform policy decisions around future hydropower projects on the Yarlung Tsangpo, where China has announced plans for additional large-scale dams including one in Medog. Whether the data gathered by the system will be shared with downstream nations — particularly India and Bangladesh — or used exclusively for domestic environmental compliance remains to be seen. The technology's scalability across other Chinese dam networks, including facilities linked to the Three Gorges Dam system, will be the next benchmark to watch.