Xiashu Tech’s AI debuts at IPPE, redefining poultry farming
The 2026 International Production & Processing Expo (IPPE) is about to open its doors, and Shanghai Xiashu Intelligent Technology Co., Ltd. (Xiashu Tech) is gearing to make a powerful impact in smart farming solutions and will unveil the company’s latest generation of AI-powered intelligent equipment.
At this year’s IPPE, Xiashu Tech will highlight 6 core products in the field of artificial intelligence and agriculture.
AI vent sexing equipment
The AI vent sexing equipment is the world’s first commercially applied sexing system for day-old chicks, integrating high-resolution imaging, multi-spectral sensing, and deep convolutional neural networks.
This device performs non-contact visual analysis on day-old chicks/ducklings. Its algorithms accurately identify subtle morphological differences in the vent, achieving high-precision sex determination. With an accuracy rate of over 99% for mainstream commercial breeds such as white-feather broilers and yellow-feather broilers, a single-machine can process up to 1,400 birds/hour. This significantly improves efficiency compared to traditional manual vent sexing and solves the problem of accuracy fluctuation caused by human fatigue and varying experience levels.
Feather sexing equipment
The feather sexing equipment utilises AI visual detection algorithms to determine sex by capturing and analysing the genetic marker of feather growth speed differences (fast/slow feathering) in chick wings.
The product features a unique single-person dual-station design for intuitive operation. A single sexing process takes about 1 second, with a maximum efficiency of 3,000 birds/hour and an accuracy rate of no less than 99%.
This technology drastically reduces the strict requirements for operators’ vision, experience, and age, significantly saving on training costs and reducing reliance on manual labour.
Intelligent egg belt counting system
Integrating high-precision sensors and image recognition technology, the intelligent egg belt counting system can count eggs on conveyor belts in real-time. It simultaneously detects abnormal eggs such as cracked, dirty, and deformed eggs, and analyses egg weight distribution, providing immediate data support for production management and nutritional adjustments.
Egg counting accuracy can reach above 99.9%. This product has been applied in Malaysia, Thailand, Indonesia, and China.
Poultry image weight estimation system
Based on deep learning vision technology, this system, which is widely applicable to common poultry species such as chickens, ducks, and turkeys, is designed for floor-reared poultry farms to estimate weight throughout the poultry’s full life cycle. It tracks behaviour trajectories and identifies sex and abnormal states in real-time, achieving high-precision weight estimation based on dynamic image analysis with a comprehensive accuracy of over 97%.
The system is further introducing time-series prediction models to forecast future weight changes based on growth curves and generate targeted feeding adjustment plans.
Farm quantity counting and behaviour monitoring system
Utilising surveillance cameras, this system counts poultry inventory and monitors flock anomalies to ensure farm health. Paired with standard surveillance cameras, the system not only precisely counts poultry but also keenly captures abnormal behaviours such as huddling and isolation.
With an accuracy rate of over 96%, it enables farmers to grasp subtle changes within the farm in real-time, facilitating timely and scientific management decisions. Widely applicable to various poultry species such as chickens and ducks, this system is currently in use across numerous farms in China and Southeast Asia.
Slaughter-end defect recognition and counting system
Based on AI technology, this system simultaneously counts live chickens and ducks at slaughterhouses, while also measuring weight, detecting defects, and performing grading with an accuracy of over 97%.
The equipment utilises robotic arms to automatically unhook birds identified with specification abnormalities based on visual detection, achieving fully unmanned detection and removal. It also provides full-process information management, helping users improve farming efficiency and precisely control slaughter weights.
We look forward to meeting you
We look forward to meeting global partners at our booth (B10046) at the upcoming IPPE. Come experience this AI-driven revolution in farming efficiency and discuss the next decade of smart livestock farming with us. Let us join hands to create a new ecosystem of digital farming!
