Font size:

Smart Poultry - Connecting Poultry Supply Chain with Smart Technology

According to the Agricultural Statistics Annual Report (109) of Ministry of Agriculture, Executive Yuan, poultry farming accounts for 47.08% of the output value of animal husbandry, of which broiler chickens account for the largest 27.51%, and the output value is as high as NT$46.4 billion, which is an important source of domestic economy. Taiwan is located in a subtropical region with high temperature and humidity. Typhoons and torrential rain are often raging in summer, and cold currents are frequent in winter. Under the general environment of open poultry houses, the production performance of poultry is greatly affected. On the other hand, open poultry houses are easy to be invaded by wild animals, which also increases the difficulty of feeding management and disease prevention, and is not conducive to automated data collection.

The poultry industry chain includes breeding, raising, slaughtering, processing and sales. Through the digitization of raising experience and the intelligent control mechanism, the standardization of raising technology and the optimization of poultry growth mode are completed, thereby reducing the infection of poultry diseases, improving the hatching rate and raising rate, and enabling the raising experience to be passed on through the data. The purpose of saving labor is to solve the problem of labor shortage in the industry. By integrating the poultry industry's digital service expert system, chicken image and voiceprint recognition and patrol robots, it supports the decision-making of poultry raising management, enhances the value of products and services, and improving the raising efficiency and animal welfare, thereby improving the business performance of the poultry industry.

In addition, by introducing the automation systems used in technology factories, such as conveyor belts, robotic arms, automated guided vehicle, RFID technology and warehouse management, etc., to assist in the automation and intelligent improvement of meat cutting and egg washing and processing, etc. It not only reduces labor burden and occupational injury of employees, but also improves operation efficiency, and has better management benefits for output and production scheduling, warehouse management, and order control.



Intelligent Continuous Incubation System for Poultry

In the past manpower was required to monitor the operation of the hatching incubator frequently during the process of hatching eggs. After introducing this system, managers can remotely monitor the temperature, humidity and ventilation of the incubator by the APP, and adjust the ventilation equipment through the control system to maintain a suitable environment. The data is sent back to the company management center in real time. This system increases the hatching rate by 4%, reduces manpower inspection during hatching, and improves the overall efficiency of hatching eggs.

Smart Breeder Selection System

The breeder selection system developed in response to the needs of domestic breeders. It uses video data to screen breeders to help the industry select breeders with good fertility and performance, and eliminate unsuitable breeders as soon as possible, which greatly reduces the cost of breeder selection. Small, easy to move, easy to sterilize and easy-to-operate selection and grading equipment, one system can be moved to multiple chicken houses, greatly reducing the cost of system construction and reducing the need for manpower and time consumption.

Individual Weighing
System for Breeding Duck

This system combines poultry foot rings with RFID tags and electronic scales to record the individual weight of breeding ducks promptly. Each individual measurement is completed in average 15 seconds. By analyzing the growth data of breeding ducks, it discloses the relationship between growth and feeding management at different ages. It also provides the growth data of individual duck. Using this system, individuals can be clearly identified, which greatly decreases repeatedly redundant sampling. The measurement records are directly digitized, reducing the worker's burden and errors during transcription. The number of workers has been reduced from three to two. This digitalizing system improves the operational efficiency and achieves labor saving in breeding poultry management.

Smart Automatic Environmental Monitoring and Control System for Poultry

This project develops an environmental monitoring and management system suitable for poultry farming based on Taiwan's climate characteristics. Various sensors in the poultry house are used to automatically detect the temperature, humidity, gas concentration (CO2), wind speed, water consumption, feed amount, poultry weight, etc. Other management data are also collected and uploaded to the management cloud platform. Using these data, the poultry growth curve and feed efficiency are established. The system also conducts the environmental control in the poultry houses according to the suitable environment for different growth stages. All of the conditions in the poultry are monitored by the system APP and webpage screen to provide real-time information. When the system detects abnormality, it automatically turns on the devices for regulation, and simultaneously issue an alarm to notify the manager. This system increases the breeding rate of poultry by 5%, alleviates the manpower requirement for daily management, and improve the management efficiency effectively.

Patrol Robot for Flat-feeding Chicken House

A Poultry Patrol Robot is a small tour-wheel-drive mobile vehicle equipped with deep-groove tires, so that it is very suitable to move on a littered poultry house. Integrated with in-door positioning system and automatic latch fence, the robot can navigate from section to section in the whole farm. While the robot is in patrol, chickens around the robot can be driven to move and exercise. In the meantime, they are likely to drink more water and eat more to improve meat production. The environmental sensors and cameras on the robot can also transmit real-time data to farmers. In conclusion, the Poultry Patrol Robot can be helpful in the daily farm routine works.

Chicken voiceprint recognition system for respiratory diseases detection

In the poultry industry, the farmers often estimate the risk of poultry respiratory diseases by the poultry sounds. This research established a chicken voiceprint recognition model by machine learning technology. It could automatically identify whether the chicken voiceprint is healthy or not, and calculate the proportion and number of abnormal chicken sounds, and upload results to cloud or server via wireless network. The accuracy of this chicken voiceprint recognition model is higher than 80% . This system could be installed in the poultry house to monitor the voiceprint of chickens at all times, and provide to farmers for assessing the chicken health condition.

Quantifying chicken dispersions and movements on commercial chicken farms using convolutional neural network and top-view monitoring system

A typical chicken farm usually contains thousands of chickens, rendering patrolling a laborious and time-consuming task. Therefore, this study employed a top-view monitoring system to acquire chicken images from a bird’s-eye view. This study proposes an approach for automatically detecting the movements and quantifying the dispersion of chickens on commercial chicken farms using top-view monitoring system and convolutional neural networks (CNN). The collected images were employed to train a CNN model to detect the positions of chickens. After the chickens were detected, the dispersion and movements can be calculated. The proposed approach achieved an average precision of 94.24% in chicken detection and an accuracy of 95.30% in chicken tracking. The quantified data can be displayed on an interactive website and can help chicken farmers detect disturbances to chickens and abnormal activities of chickens.

The analysis and prediction system for egg laying performance in egg laying farms

The poultry production management expert system integrates the data of poultry house environment, poultry raising, bird body image and voiceprint to form a basic information database. A predictive decision-making database built by an AI model can cooperate with the mobile device APP interactive chat robot mode to provide the producers with environmental forecast notification, feeding management early warning, industrial word cloud information, online consultation and other services. In environmental forecasting, the predicted R2 reaches 0.89. In feeding management, the predicted R2 of water consumption reaches 0.86, and the predicted R2 of sick and dead birds reaches 0.78. The real-time situation of the poultry house and the condition of the poultry could be notified by using this system, and the corresponding treatment suggestion in advance could be obtained before the abnormal situation occurs. The remote monitoring and elimination of abnormal conditions could be operated via the digital interface, which not only simplifies the management process, but also increases the management efficiency. The breeding rate and feeding efficiency of poultry could be raised as well.

The prediction and analysis system for egg laying performance in egg inductry

The egg industry encounters the threats of labor aging and shortage, climate change and of avian Influenza. The managers have no choice but to upgrade their facilities or management method to overcome these threats. This project intended to analyze the daily management and production data to establish an AI model to predict the egg laying performance. The R2 value is 0.94 and error is 1% using this model to predict the egg production rate five days later. This technology provides production prediction in real time and helps the managers to make effective decision for the egg industry.

Chicken automatic cutting equipment

Due to the large number of size of native chicken, it is difficult to directly introduce automatic cutting equipment for white broilers. Therefore, a lot of manpower is required in the cutting process. This project developed automatic chicken cutting equipment, which has six types of cutting knife program options (customizable expansion is possible), and different cutting times and cutting distances can be set. The machine also has the function of Internet of Things, and the operation status of the machine can be transmitted to the monitoring computer. The automatic cutting equipment takes about 10 seconds, which can greatly improve the operation efficiency and improve the labor of manpower operation timing rotation.

Flexible Transport /Palletizing System for Egg Packaging Factory

This project develops a flexible transport /palletizing system based on a robotic arm combined with radio frequency technology (RFID), which has the functions of carton horizontal transport, vertical palletizing, abnormal product detection and production records. This system handles about 60% of the product in the egg packaging factory, saves 25% of the working time, and greatly reduces the physical load and occupational injuries of laborers, and also reduces the probability of packaging errors.

Intelligent Storage System of Processed Duck Eggs

The intelligent storage system of processed duck eggs can use automatic guided vehicles (AGV) to unload and transport pickling buckets, establish automatic storage positioning and designated delivery functions, and integrate with the warehouse management system. The warehouse management system can automate the inventory records, complete the production history and inventory, and set the warning function at a low or high level. It is the first intelligent production management system for duck egg production and processing. It can use the data analysis function to provide order and delivery management and achieve the benefits of saving labor costs.