The breeding industry ushered in a breakthrough in transformation and upgrading


  AI pig raising is by no means overkill

  Our reporter Li He

Based on big data analysis of eating conditions, smart equipment can achieve precise feeding; electronic ear tags quickly detect the physical condition of livestock and poultry; environmental detection systems monitor temperature and humidity, hydrogen sulfide and other harmful gas concentrations, light intensity and other environmental factors in real time... these control systems Both can be integrated into a mobile phone, allowing farm managers to obtain abnormal information in time, realize remote precise control, and control related equipment based on the monitoring results.

  At the 18th (2020) China Animal Husbandry Expo, which opened on September 5, the breeding "black technology" made people shine.

It can automatically and accurately feed the sows, piglets, nursery pigs, and finisher pigs at different growth stages and different breed characteristics; electronic temperature measurement ear tags can quickly detect the physical condition of the pigs; according to the temperature and humidity in the house, air quality and other parameters, the system can automatically feed Alarm and remotely turn on the fan to adjust the parameters... Many traditional pig farms are accelerating the transition to “digital, intelligent, and informatized”. Livestock and poultry breeding has not only become more scientific and technological, but also improved production efficiency.

  "The aquaculture industry is labor-intensive, and the workplace environment is relatively harsh, and there may even be odors and odors. At present, the structure of my country's labor force is changing, and young people are unwilling to work in farms, and labor costs continue to increase, so intelligent Pig raising is an inevitable trend for the development of large-scale pig farms in the future." On September 20, Hai Kumamoto, a second-level researcher at the Beijing Institute of Animal Husbandry and Veterinary Medicine, Chinese Academy of Agricultural Sciences, said in an exclusive interview with a reporter from Science and Technology Daily.

  Realize precise feeding based on big data analysis

  Feed costs account for 60%-70% of pig farm breeding costs. Feed intake is closely related to pig weight and health status.

For example, pregnant sows need to obtain sufficient nutrients from the diet to meet the growth and development of the fetus and reserve some nutrients for the lactation period; lactating sows must ensure the maximum feed intake and not excessive, ensure good body condition and good breast milk , Thereby shortening the estrus interval.

  Kumamoto said that based on big data analysis and algorithms, researchers have built a pig's feed intake prediction model at various stages, and realized the dynamic and precise analysis of pigs' feed nutrient requirements.

Through the integration of radio frequency identification (RFID), perception of feeding behavior, precise control technology and equipment for feed cutting, and linkage with feed intake models, a personalized intelligent control system for the feeding of pregnant sows has been developed, which has achieved the Pigs provide customized feeding programs.

  According to reports, RFID is a kind of automatic identification technology. Non-contact two-way data communication through RFID can achieve identification goals and data exchange.

Today, this technology is widely used in the smart pig industry, mainly for individual identification.

  Practical verification has shown that through the intelligent control system for precise feeding of pregnant sows, the remaining feed ratio after the sow eats is less than 1%, and the feed intake ratio is greater than 95%, which greatly improves the feeding efficiency.

  "A precision feeding smart machine can feed 40 pigs. In order to avoid multiple pigs grabbing food, pigs are generally allowed to take turns to enter the smart machine to eat. But pigs are alive after all, and some pigs will not quickly eat If you leave, the pigs behind cannot enter, which reduces the feeding efficiency of the pigs." Kumamoto said that by combining artificial intelligence technology with the habits of pigs, using visual cognition, free intake and automatic feed intake recording system, The intelligent control system can actively analyze. Once it is judged that the pig has reached the predetermined food intake and entered by mistake, the door of the feeding trough will automatically open to allow the rear pigs to enter and the front pigs to arch out to improve equipment utilization efficiency.

  At the suckling pig feeding station, a group of piglets less than 21 days old are forming a circle to feed themselves.

Feeding 6 times a day can not only maximize the feed intake of lactating sows, but also provide sufficient milk to the piglets, protect the weight and immunity of the piglets, and shorten the lactation period from 28 days to 21 days. Allowing the piglets to meet the requirements for early weaning weights can also enable the breeding sows to enter the next estrus and production cycle after a short adjustment, and ultimately increase the sow's productivity.

"According to the different physiology or growth node conditions of each pig, through the dynamic prediction model of the diet intake, the nutritional control measures that each pig should take at different stages are intelligently determined, whether to choose controlled feeding, measures to promote feeding or Free-feeding avoids the impact of human factors on pig production. The ultimate goal of these measures is to reduce feed waste, realize the maximum production potential of pigs, and reduce the production cost of pig farms." Kumamoto said.

  Electronic ear tag real-time monitoring of pig body temperature

  In 2019, when the African swine fever broke out, a large number of sick pigs were buried and killed, and pig farmers suffered heavy losses.

Electronic ear tags were used to trace the origin of pigs.

"Today's electronic ear tag chips are loaded with sensitive temperature measurement circuits, which can obtain pig body temperature data in real time, predict the physical condition of the pigs, and quickly identify the pigs that may be sick, which helps farmers to take timely isolation and other measures. Significantly reduce the probability of disease transmission in pig farms." Kumamoto Kai said.

  The electronic ear tag is embedded on the pig's ear, and the pig ear is closely connected with the intelligent sensing device. Therefore, the pig's body temperature can be monitored in real time, and the temperature measurement error is small.

The ear tag will also automatically send the pig's body temperature data to the background according to the set time, such as every 10 minutes.

Once the pig's body temperature is abnormal, the LED light on the ear tag will immediately alarm and report to the cloud server at the same time, prompting the farm manager to take timely isolation or other measures to achieve efficient epidemic warning.

  Whether an animal is sick or not, in addition to changes in body temperature, food intake and activity levels may also decrease.

The electronic ear tag can also be combined with the precision feeding intelligent control system and video to monitor the status of each pig in real time.

And through the big data obtained, the management of livestock archives, traceability, group management, growth status and distribution data summary management is realized.

  Environmental factors such as temperature and humidity in the pig farm, the concentration of harmful gases such as hydrogen sulfide, and light intensity have a greater impact on the growth rate and development of pigs.

A good and comfortable environment not only promotes the growth and development of pigs, but also has a relatively high productivity and feed conversion rate.

In the past, the omission of environmental management in pig farms often caused pig infections.

  "Environmental control is an important aspect of AI pig raising." Kumamoto said.

  The artificial intelligence system in the pig farm environment collects online air temperature and humidity, carbon dioxide, ammonia and hydrogen sulfide concentration and other environmental information of the farm through smart sensors, and uploads the information using the Internet of Things.

The control system intelligently judges whether the pig farm environment is suitable based on big data such as the area of ​​the pig farm and the density of pigs, and takes corresponding measures according to the specific situation.

For example, if the concentration of carbon dioxide or ammonia in the air of a pig house is too high, the intelligent environmental control system will automatically turn on the fan to increase the air exchange; if the temperature is too high or the humidity is too low, it will increase the water volume of the indoor water curtain. The role of cooling and humidification.

At the same time, the system will collect data, analyze the response of animal behavior to environmental comfort, establish parameter models and thresholds for judging comprehensive environmental comfort, and analyze and establish the relationship between environmental parameters and feed conversion rate and production performance.

  The next step is to improve accuracy and reduce equipment costs

  Kumamoto Kai said that whether it is a precision feeding smart machine or a smart environmental control system, its control system can be integrated into a mobile phone. The farm manager can obtain abnormal alarm information in time, realize remote precise control, and according to the monitoring results, Intelligently control related equipment to achieve the goals of healthy breeding, cost saving, labor reduction, energy saving and consumption reduction.

  Artificial intelligence pig raising has the advantages of improving production efficiency and reducing costs.

For example, the "Pig Precision Feeding Management Theory and Intelligent Equipment Creation" project jointly completed by the Beijing Institute of Animal Husbandry and Veterinary Medicine of the Chinese Academy of Agricultural Sciences and Beijing Agricultural College has created 7 types of intelligent feeding equipment from breeding pigs to commercial pigs; 152 sets The pig feed nutrient prediction model has a prediction accuracy of 93%; 38 sets of main nutrient prediction models at different physiological growth stages of pigs have been created; electronic identification, dynamic weighing, etc. are integrated to realize the automation and intelligence of the performance determination of breeding pigs.

  Among them, the application results of the intelligent feeding system in pig farms show that it can reduce feed waste by 7%-10%, increase feed conversion efficiency by more than 5%, and save labor by more than 35%.

  Kumamoto Hai said that at present, the development of artificial intelligence in the aquaculture industry is in its infancy, and there are still some problems and deficiencies.

Its accuracy needs to continue to improve. For example, using smart collars to monitor the rumination of cows. Because cows’ necks are of different sizes and will keep moving, the collars around the cow’s necks sometimes move away from the monitoring position. The accuracy is about 95%, which is far behind the accuracy of foreign advanced SCR electronic collars.

  Secondly, the price of smart equipment is higher than that of similar breeding equipment, which can only be afforded by large farms.

However, with the development of artificial intelligence, more research and development institutions and enterprises will invest in the research and development, production and application of artificial intelligence technology and equipment, and the cost of smart equipment will drop accordingly.

  "In artificial intelligence breeding, single-point technologies and systems cannot solve the efficiency problem of the pig industry. These technologies, facilities, equipment, and information need to be integrated. For example, the implementation of artificial intelligence needs to be considered in the planning and design of pig houses. Under the circumstances, the integration of top-level breeding process design and operation is realized. Through the use of artificial intelligence, not only can the overall efficiency of the pig breeding industry be improved, but also animal welfare and biological safety can be guaranteed, and ultimately the effective supply of livestock products." Kumamoto said.