Long queues were seen at Shenzhen University's (SZU's) campus canteen last month where teachers and students waited to try a free "AI chicken soup" cooked with black silkie chickens that were gifted by thankful Guizhou farmers.
A dozen students from SZU's Tencent Cloud and AI Class developed an AI-based farming system for black silkie chickens last year.
A Shenzhen University student checks chickens at a chicken farm in Chishui, Guizhou Province. A dozen students from the university's Tencent Cloud and AI Class developed an AI-based farming system to help the farms digitalize the management of the birds and raise production last year. Courtesy of Shenzhen University
The Tencent Cloud-based system took eight months of designing and testing before it was launched at a chicken farm in Chishui, Guizhou Province. The system helps farm workers detect inactive chickens that could be potentially unhealthy and reports attacks from weasels or other wild animals in real time. These preemptive measures not only guarded against epidemics and loss of chickens from other causes, but also helped the farm produce about 60,000 more chickens in a six-month period, 30% more than previously.
AI empowers farming
Wang Yifeng, an SZU student who participated in the Chishui project, had already gained some experience during a project in 2022, in which he and 15 classmates designed a miniprogram to help locals better manage goose farms in Shantou, Guangdong Province.
The goose farm project resulted in Wang and his classmates winning a provincial-level gold medal in the ninth China College Students "Internet+" Innovation and Entrepreneurship competition. When city authorities from Chishui expressed their interest in upgrading their traditional chicken farming methods, Wang and his team were at the forefront and ready to work.
Starting in April 2023, Wang and his teammates went on multiple field trips in Chisui's dense mountainous forests. For over 300 years, local farmers have raised their prized black silkie chickens in the primal forests.
According to Wang, locating and tracking 250,000 chickens was the first step to digitalize management. The task was challenging because computerized visual detection in natural environments faces unpredictable factors that can affect accuracy.
To solve the problem, the students consulted a dozen published essays. After numerous trial and error cycles, they finally designed a foot ring that can track the chickens, count their steps, take their body temperatures, and detect abnormal gait. The students also deployed monitors in chicken coops to ensure that the animals have a healthy and comfortable environment.
Then, using an algorithm to analyze the collected data, the students also designed a WeChat miniprogram app and an intelligent terminal that allows the farmers to monitor the chickens and alerts them to potential hazards. The app also allows consumers to purchase chicken products. The miniprogram enables the staff to post videos of the free-range chickens and livestream their activities as they work in the mountainous environment, kindling an interest in consumers and boosting sales.
In eight months, all 250,000 chickens had been identified and their data was logged in the system. Over 100,000 pictures were also fed into the AI to teach it to recognize abnormalities in the chickens.
Chickens raised 'on the cloud'
Chen Guofan, head teacher of the Tencent Cloud and AI Class, beamed with pride for his students as he ate the hot chicken soup. "I fed them enough chicken soup for the soul, and now they repay my support with yummy real chicken soup," he joked.
Technological progress greatly reduced the students' workload when they embarked on the chicken farming project. With assistance from Tencent technicians, the team used tools offered on the Tencent Cloud TI Platform to identify chickens and input data semi-automatically — the process led to a 40% to 60% reduction in total workload. AI model training, deployment, and software development were all completed on the platform.
According to Wang Cairong, a Tencent employee in charge of cooperative projects involving higher education institutions, his company offered resources and technological assistance in database building, cloud storage, server management, and AI model training.
Market-oriented education
The Tencent Cloud and AI Class, first initiated in 2018, is a strategic partnership between Tencent and SZU designed to cultivate urgently needed talent in big data, cloud computing, and AI sectors.
The first batch of its graduates were met with high levels of enthusiasm when they entered the job market in 2022. More than half of the graduates found positions at tech giants such as Tencent and roughly 40% enrolled in postgraduate programs at universities in China and overseas. While many Chinese college graduates are scrambling for jobs, the success of these students demonstrates the practicality of higher education pathways that give students the opportunity to gain hands-on expertise and solve real-world problems.
Huang Hui, dean of Shenzhen University's College of Computer Science and Software Engineering, said the AI-assisted chicken farming project is a paradigm of effective education. "The project gave our students the opportunity to tap into their creativity to solve complicated engineering problems in the real world, helping to transform them into competent AI engineers needed by the industry," Huang said.
Zhang Xiaohong, vice president of SZU, confirmed the significance of the project. "The close cooperation between our university, Tencent, and the Guizhou farmers gave our students the chance to better understand the AI industry, to better master theories and technologies, and to improve their communication and problem-solving skills. Learning in practice is the best way to learn," she concluded.