Voice of Money, Rainforest
A group of cute little monkeys live in the Costa Rican rain forest. They have slender limbs, a dark brown body, and round and small heads. Their adult body length is about 60 cm. When they hang on the tree, they are like a spider, so they are called spider monkeys.
Spider monkey reproduction is very similar to humans.
Female monkeys give birth to a pup eight months after conception. The pup's growth is very slow, and it takes about two years to completely wean. Under the shelter of the big spider monkey, the little spider monkey can thrive.
However, in recent decades, due to the continuous destruction of their habitats, the threat of deforestation, reduced precipitation, tourism development, and predation by indigenous people, the number is rapidly decreasing.
Spider monkeys were included in the World Conservation Union's 2008 Red List of Endangered Species ver 3.1.
Rainforest protection, Voice brings a silver lining
The unexpected turnaround occurred in 2011. Engineer Topher White stumbled upon an illegal feller while on holiday in Borneo, and the felled trees were only 100 meters away from the location of the ranger station.
The sounds in the tropical rainforest are diverse and complex
It is difficult for humans to distinguish the noise of the chainsaw felling
Topher is determined to start a charity organization, looking for solutions to protect the rainforest. Through unremitting efforts and donations from the society, Rainforest Connection (RFCx) Rainforest Protection Organization was established.
By upgrading a large number of recycled old mobile phones into solar rainforest monitoring equipment, they are scattered in the depths of the jungle.
In this way, no matter the downpour, the sun is shining, or the weather is wet all day, once the monitoring system finds abnormal sounds of theft, such as the sound of a chainsaw or truck, it will alert the first time and push the specific location of theft to the local rangers to help them quickly search.
Topher said that due to the reliability of Huawei's mobile phones, rainforest protection organizations have also used Huawei's mobile phones extensively in Costa Rica's rainforest monitoring. The range of land protected by Huawei mobile phones has exceeded 2500 square kilometers of land (about 200,000 football fields).
This innovative solution faces multiple challenges:
1. How to collect and transmit audio data in the harsh environment of high temperature, high humidity, and no fixed power supply, and store and manage these continuously growing huge data on the back-end platform.
2. How to conduct real-time data analysis and accurately determine the location of the illegal logging. The sounds in the rain forest are complex and complicated, coupled with the huge amount of data, the algorithm for precise identification is very strict
Huawei and the rainforest protection organization are carrying out a series of cooperation to develop an innovative platform that includes collection equipment, storage services, and intelligent analysis to jointly protect the tropical rainforest ecology.
Use Huawei Cloud's big data capabilities to store and manage audio data collected at various collection points. Based on the artificial intelligence service HUAWEI CLOUD AI and ModelArts tools, develop intelligent algorithm models to achieve accurate recognition of chainsaw and truck noise.
Huawei technicians are testing the collected spider monkey sounds on the ModelArts platform
In addition to preventing logging, rainforest protection organizations also need to monitor the survival status of endangered animals such as spider monkeys.
To this end, Huawei and the Rainforest Protection Organization have jointly built a spider monkey language intelligence analysis model to provide information about habitat and living habits. The data collected by mobile phones is not only used to protect spider monkeys but also enables the protection of countless species.
Jenna Lawson, a PhD student researcher at Imperial College London, pointed out: "AI allows me to train machines and algorithms to help me detect species. Now, I am trying to collect more than 200,000 data. Based on these data, I can find the calls of different animals, and then make a map of the habitat distribution of different species. "