Hung, Chia Hung（China Productivity Center Smart Agriculture Promotion Department）
Along with diminishing working population and significant difficulty in recruiting workforce, Taiwan’s conventional agricultural industries have been faced with increasingly short supply of talents. Particularly for pig raising industry in Taiwan, aging of farmers has led to shortage of needed labor and increases in labor cost have posed challenges as well. Besides the labor shortage, the low breeding rate has resulted in low profitability of the pig raising industry. The pig production cost in Taiwan is 20-70% higher than that in the US and European countries, and this is because pig farms in Taiwan are too small in size to create economies of scale and the imported feeds incur high cost.
The high piglet breeding cost is another reason for these problems. The average pig raising rate of 70% in Taiwan is significantly lower than those in the US and European countries. In view of such critical issues, Taiwan-based Wan-Chi-Sheng Agricultural Technology Co., Ltd. (“Wan-Chi-Sheng Co.”) has thought over application of information technology to automated and smart pig raising in a bid to reduce cost and expenses incurred and labor input as well as hike overall piglet breeding rates.
Technological Upgrading Transforms Pig Raising Industry
According to statistics released by Ministry of Agriculture (formerly Council of Agriculture), there were 6,133 pig farms with 5.375 million pigs raised in total, including about 590,000 reproductive sows, in Taiwan in 2022. As conventional pig farms rely on labor to breed and nurse piglets as well as regularly patrol pig stalls to prevent sows’ overlying piglets to crush them to death so as to maintain piglets’ survival rates at as high levels as possible. However, the increasing labor cost renders it necessary to seek new methods of preventing sows from crushing piglets.
Through taking part in the “husbandry industry’s participation in smart agriculture” project in 2022, Wan-Chi-Sheng Co. obtained subsidies to develop a smart image recognition-based system of monitoring sows and piglets living in farrowing stalls. The system is an early warning one that, through smart recognition of monitoring images, can detect conditions posing threats to piglets’ survival and then send warning to management staffs for appropriate handling in order to maximize piglets’ breeding rates.
Wan-Chi-Sheng Co., after participating in the project, proceeded with the application research. By multiple-facet exploration of relevant knowledge and advanced technologies, the company has developed the early warning system through integrating image capture, IoT, AI, cloud computing and other technologies. The system is mobile and can undertake 24-hour recording of images of sows and piglets in every farrowing stall.
For the system, an AI algorithm is used to recognize captured images of key behaviors, including sows’ postures, drinking water and eating feeds, lactation as well as piglets’ activities to analyze sows’ breeding behavioral models and piglets’ action models. With camera lenses in place of human eyes, the system preliminarily analyzes image data to provide information and data on pigs for use as reference in future pig farm management.
Specific to possibly threatening conditions which are not found by management staffs, the system can send early warning for them to immediately respond, thereby reducing use of manpower and hiking breeding efficiency.
According to the evaluation of Wan-Chi-Sheng Co., application of the system is expected to hike piglet breeding by 10%. Based on the operational scale of the experimental pig farm and its general breeding rate of about 80%, the hike in piglet breeding is equivalent to an increase in the total number of grown (adult) pigs by 22,288 a year or annual production value by NT$20.59 million. The fruition makes Wan-Chi-Sheng Co. confident of promoting application of the system to all pig farms to create a large commercial value in the future.
due to increasingly intense competition in the international food market and rising prices for materials of feeds, pig raising cost in Taiwan is on the rise and accordingly profitability from pig raising becomes limited despite high sales prices for pigs. Consequently, considerable profitability hinges on keeping improving pig raising performance and decreasing labor cost incurred.
Data-based management is key to overall upgrading of pig raising industry in Taiwan. The case of Wan-Chi-Sheng Co. demonstrates that introduction of various sensors, automatic image capturing devices and nearly standardized operation can improve quality of image data as well as collect more real-time and precise images to reflect real raising conditions at pig farms.
Objective data enables analyses to prove methods of entire pig raising management to thereby hike production capacity and ease shortage of husbandry manpower. Facing globalization competition in the future, Taiwan’s pig raising industry needs to strengthen capabilities, including prevailing use of data-based management and automated equipment as well as integration of software and hardware, in a bid to transform and upgrade operation in line with movement toward an era of intelligence.