#China #Huawei #Xiamen #Chinese_White_Dolphin #AI_Tech

Innovative AI Solution Achieves Early Success in Protecting Chinese White Dolphins

In a significant achievement for marine conservation, Huawei and its collaborators have unveiled promising preliminary results from an innovative AI-powered solution designed to explore and protect the endangered Chinese white dolphins in Xiamen Bay. Launched three months ago as part of the Huawei TECH4ALL initiative, this groundbreaking project aims not only at preserving this iconic species but also at utilizing advanced technology for environmental protection.

The initiative has achieved remarkable milestones, including the identification of 13 individual dolphins from a comprehensive dataset comprising 2,820 images and videos. Impressively, the system has reached an individual identification accuracy exceeding 90%, alongside an 85% detection rate for complex behaviors. Moreover, data labeling efficiency has improved by 400%, showcasing the substantial impact of AI in enhancing research processes.

Addressing the Threats

Chinese white dolphins are a nationally protected species and are classified as endangered on the IUCN Red List. The Xiamen Bay, which is home to 51% of China's dolphin population, faces significant threats from shipping activities, fishing, and coastal construction. Noise pollution, entanglement in fishing nets, and habitat loss further jeopardize their survival.

Cui Yangyang, director of Huawei's TECH4ALL program office, notes, “The data-driven insights generated by AI assist conservationists in formulating targeted protection measures to understand and respond to the threats these dolphins face, ensuring they can survive and thrive in the wild.”

The project involves collaborations with the Third Institute of Oceanography under the Ministry of Natural Resources of China and China Mobile. It features a unique identification system that utilizes the distinct markings on the dolphins' dorsal fins to identify individual animals.

Wang Xianyan, leader of the endangered marine species research and conservation team at the Third Institute, emphasizes the efficiency brought by AI, saying, “Surveys are now more efficient, and protection measures are more precise.” The data provided by the AI recognition system forms an essential foundation for future conservation strategies based on individual survival rates, reproductive dynamics, and social interactions.

Enhanced Monitoring Through Technology

The AI solution incorporates multiple technological components, including image preprocessing, AI inference and detection, cropping of fin images, data classification, and cloud display. Each dolphin has its own file, enabling researchers to monitor the status of individual animals. Long-term data collection is critical for formulating effective conservation strategies, allowing researchers to identify population counts, distribution, age structure, reproductive behavior, and the various threats dolphins encounter.

Previously, manual monitoring was labor-intensive and yielded unreliable results necessary for tracking population dynamics and developing precise conservation actions. According to the Third Institute of Oceanography, the next 10 to 15 years will be crucial for the population's growth.

The role of the Chinese white dolphin extends beyond ecological significance; they are crucial participants in coastal ecosystems and carbon cycling processes. Feeding primarily on fish that consume plankton, the dolphins play an indirect role in the carbon cycle as plankton absorbs CO2 and sequesters carbon through photosynthesis. Thus, safeguarding the dolphin population and ensuring the integrity of ocean ecosystems are vital for combatting climate change.

In addition to AI-driven analytics, a 5G-A network encompassing 10 base stations covers an impressive 330 square kilometers of the bay, facilitating real-time monitoring of key conservation areas. By utilizing the integrated sensor and communication capabilities of 5G-A and data from radar systems, visual terminals, satellites, and vessel tracking through the Automatic Identification System (AIS), the movement of ships can be monitored up to 20 kilometers away.

The system issues AI-supported alerts within seconds for violations such as boundary breaches or speed limit offenses, allowing law enforcement authorities to swiftly locate offending vessels based on land surveillance. To date, 12 ships have been investigated for potential threats to the dolphins.

Thanks to continuous research efforts coupled with AI utilization, the elusive habits of the Chinese white dolphins in Xiamen Bay are gradually being revealed, sparking hope for their continued survival. This innovative approach exemplifies how technology can intersect with wildlife conservation, paving the way for more sustainable interactions with our marine environments.