Transportation Machinery and Feeding Systems for Pigs in Multi-Storey Buildings: A Review

Abstract

In recent years, in order to save land resources and ensure biosecurity, multi-storey buildings for pig rearing are replacing traditional flat-floor methods in pig farms domestically and internationally. However, the precision, stability, and adaptability of pig feed transportation in multi-storey buildings have brought certain challenges to the development of large-scale pig raising. The uniqueness of this paper lies in the comprehensive review of recent research advances in various transportation machinery and feeding systems from the perspective of both pig feed transportation and feeding systems compared to other papers that singularly present pigs’ feed transportation machinery or feeding systems. In addition, this paper provides an outlook on the potential for coupling power sources for pig feed transportation and pig transportation, providing insights for future research and development. First, the paper comprehensively reviews feed transportation machinery in multi-storey pig rearing, highlighting their advantages and challenges. Then, it explores the commonly used feeding systems in large-scale pig rearing and their limitations. Finally, the paper summarizes the current issues in pig feed transportation in multi-storey buildings and examines future development trends.

1. Introduction

In recent years, with the boom in socio-economic development, the traditional flat-floor farming methods of the global pig industry have gradually failed to meet the new age of large-scale farming [1]. Multi-storey building pig rearing is a new term known by everyone in the pig industry [2]. Multi-storey building pig rearing addresses the increasing scarcity of land resources for farming. On the other hand, it helps promote the development of mechanized and automated equipment, which reduces labor costs and improves production efficiency. In China, pig farming faces the dual challenges of scarce land for farming and low mechanization. Land for farming is relatively limited due to land resources and geographical constraints. At the same time, the livestock sector’s mechanization level is only about 33 percent, with a significant reliance on manual feeding, resulting in higher production costs [3,4]. As early as the 1970s, multi-storey buildings for pigs rearing appeared in China, and at that time, the design was only a simple stacking of flat-floor pig farms, without sufficient consideration of pig trans-shipment routes, staff movement, and feed transportation. This results in a design that increases the cost of inputs and the risk of disease transmission compared to a flat-floor farm [5]. With the continuous maturity of the pig management system, multi-storey building pig rearing has been developed from the simple stacking of flat farms to a three-dimensional multi-storey building pig rearing mode adapted to the production reality [6].

Compared to flat-floor pig farming, pig farming in buildings requires much less land and labor costs. Compare the example of pig raising in Guizhou Fuziyuan building and pig raising in Zhenfeng flat-floor. It can be seen that there are significant reductions in floor space, utility costs, and total labor for raising pigs in buildings [7].

The current low level of overall mechanization in China is not conducive to further developing the modern pig farming industry [8]. Problems with feed transportation machinery include high energy consumption, clogging of feed, short service life due to wear and tear of bends, low adaptability to different building heights and layouts, and untimely troubleshooting and maintenance of transportation machinery at a later stage. The main problems in the feeding systems are poor feeding accuracy, the low accuracy of pig weight estimation, inability to judge the health status of the pig when feeding, and the complexity of the liquid feeding process.

In response to the problems mentioned above, this paper first outlines the existing feed transportation methods, such as pneumatic conveying, scraper pipeline conveying, screw conveying, and rail conveying, and lists the research on these modes of transportation in recent years. Then, these modes of transportation are compared and contrasted, pointing out the advantages and disadvantages of each. Secondly, it summarizes the feeding systems commonly used in multi-storey buildings, such as the “Gestalt” feeding system, the Velos feeding system, the fattening pig partitioned feeding system, and the liquid feed intelligent feeding system. It lists the research advances of recent years and analyzes their current problems and challenges. Finally, the challenges and future trends of feed transportation and feeding systems are discussed.

2. Feed Transportation Machinery

Feed transportation is one of the crucial links in building pig rearing, and the efficiency and stability primarily affect the economic efficiency and biosecurity level of building pig rearing [9]. Traditional feed delivery methods are mainly based on manual transportation, but manual feed delivery has problems such as contamination, cross-infection, and uneven nutrition. Most multi-storey pig farms today use mechanized transport equipment to deliver feed to reduce labor costs and increase the biosecurity level associated with traditional manual breeding.

International pig feeds are mainly in pelletized, powdered solid, and liquid form [10]. Pelletized and ground solid feeds have the advantages of easy storage and uniform nutrient content and are widely used feed types in the pig industry. The main conveying methods for pelletized and powdered solid feeds are pipeline transportation and rail transportation, and for pipeline transportation, there is pneumatic conveying, scraper pipeline conveying, and screw conveying. Although liquid feeding systems have been introduced to the Chinese pig industry, there are still problems of poor management, immature technology, and microbial contamination in liquid feeding systems [11]. Therefore, this section focuses on conveyance methods for solid feeds.

2.1. Pneumatic Conveying

Pneumatic conveying utilizes high-speed air generated from a blower to convey solid and powdered pig feed in a pipeline, mainly composed of a motor, a blower, a pipeline, and auxiliary equipment. The key to effective system operation is accurately transporting the material into the pipeline and the discharge device [12]. With the continuous development of mechanical transportation technology, pneumatic conveying as a kind of airflow as a power source of technology is gradually being widely applied in the large-scale breeding industry [13]. Compared with scraper pipeline conveying and screw conveying, a pneumatic conveying device has a simple structure and high space adaptability, which can realize long-distance, multi-directional, and multi-angle transportation. However, the pneumatic conveying method has problems of high energy consumption, easy abrasion and crushing of feed particles, and the erosion of bends. Experts have conducted some research to address these problems.

To overcome the high energy consumption of pneumatic conveying, Baker et al. provided a method for online optimization of the system to obtain the best operating point by probing for the optimal gas volume flow rate as well as the optimal gas flow rate from an isobaric pressure drop inlet, thereby reducing the energy required by the equipment and optimizing the performance of the pneumatic conveying system [14]. However, when selecting the optimal working point, only the solid flow rate was mentioned, and the shape and size of the feed particles were not mentioned. Therefore, Suetina et al. developed an adaptive regulation system for pneumatic conveying using the movement pattern, shape, and size of the material particles as the key parameters, and by investigating the effect of the key parameters on the conveyor, the pneumatic conveyor was made to work closer to the resistance boundary at the same load value, thus reducing the energy consumption [15]. However, the study by Suetina et al. did not mention whether the shape of the delivery pipe affected their study. Therefore, Song et al. investigated and constructed the correlation between energy consumption and the solid-to-gas ratio of the pneumatic conveying system by using shrimp feed particles as an example and the conveying characteristics of feed particles in long vertical pipes with elbows at both ends, V-pipes, and H-pipes by a coupled CFD-DEM method [16]. It was found that a solid-to-gas ratio of 3.69 is better for shrimp feed pellets with a solid loading rate of 0.54 kg/s to minimize the system’s power consumption. However, the fact that the pneumatic conveying system was more effective for shrimp feed pellets at a solid-to-gas ratio of 3.69 does not mean that this solid-to-gas ratio is applicable to other feeds. For this question, Freitas et al. explored the reasonable conveying conditions for pneumatic conveying of milling powder using the Zenz diagram and PLC control technology [17,18]. Among them, the Zenz diagram can provide some help and reference for exploring the solid–air ratio of different pig feeds, which can help save energy.

For feed pellets prone to crushing, Kong et al. simulated and modeled the crushing characteristics of feed particles in pneumatic conveying by CFD-DEM [19]. It was found that at high gas velocities and small bending radii, the particles break up easily, and the energy consumption is high. Aiming to address the problem of feed particles prone to breakage during conveying, Ghafori et al. proposed a novel elbow design with auxiliary air. By doing experiments, it was verified that the new elbow has the advantages of smaller average air pressure, less collision between particles and pipe wall, and lower particle fragmentation rate at 45° compared with 60° and 90° [20]. However, the manufacturing cost and longevity of the elbow have not been practically proven in the industry. To address the erosion of bends, Mills D et al. and Levy A V et al. explored the erosion rate of particles on elbows through the size, shape, and angle of transported particles, respectively, and found that the smaller the particles, the faster the erosion rate on elbows, and the erosion rate of angular particles was much larger than that of round particles [21,22].

To minimize damage to bends in pneumatic conveying by particle erosion, Duarte C A R et al. compared the degree of damage caused by particle erosion of plugged tee elbows, volute elbows, and standard elbows under the same operating conditions. The results showed that the volute elbow has superior erosion resistance under higher mass loading conditions [23]. Although, according to this study, we know the hazards of the erosion of elbows by differently shaped particles, it is difficult to know the wear of elbows in real time in real life, so Yasin Alkassar et al. modeled the empirical relationship of the erosion and wear of elbows by establishing an empirical relationship [24]. Although Ysssin Alkassar revealed a link between elbow wear and particle transport, it did not go deeper into the angle of the elbow. Zhou F et al. simulated the elbow erosion phenomenon by the Euler–Lagrange method and discrete element method using coal particle transportation. The study results showed that when the angle of the elbow was 30°, the erosion hazard of particles on the elbow was minimized [25]. The study revealed a correlation between erosive wear and particle transport, and the model’s prediction of elbow mass reduction suffers from an error of about 15% with actual experimental data. In order to extend the service life of the elbow, Guo Zihan et al. proposed three different shapes of ribs, namely, quadrilateral, isosceles trapezoidal, and isosceles triangular, to be installed in the direction of the outer diameter of the elbow. They simulated the corrosion resistance of the ribbed pipe using the CFD-DPM method [26]. The study results showed that isosceles triangular ribs have the best corrosion resistance characteristics.

Additionally, based on the surface structure of the back of the desert scorpion, Guo Z et al. constructed bionic transverse grooves, bionic ribs, and single-rib elbows with different cross-sections and conducted comparative erosion resistance tests [27]. The triangular single rib was found to have the best erosion resistance, consistent with the conclusion of Guo Z. In addition, Saluja G. et al. developed an empirical model with high generalization ability and low relative error for minimum transport boundaries to provide minimum boundary conditions for transporting powdered feeds. This new model has significant advantages in adaptability and accuracy over existing prediction models [28].

These studies provide guidelines and references for solving the problems of high energy consumption, feed particle abrasion, crushing, and bend erosion in pneumatic conveying. Pressure drops during pneumatic conveying, the solid-to-gas ratio of feed particles, gas velocity, and bends were studied to reduce the system’s energy consumption, wear of feed particles, granule breakage rate, and mitigation of bend erosion.

2.2. Scraper Pipeline Conveying

Scraper pipeline conveying is known as plug line conveying, and this subsection uses the term stopper line conveying to describe this conveying. Plug line conveying utilizes the friction between feed particles and the pressure the plug tray provides to transport the feed through the pipe [29]. The space layout of this conveying method is highly flexible, and it can be combined with multiple lines according to the actual demand or combined with other conveying methods to form a complex feed conveying system, realizing quantitative and fixed-point conveying [30]. The use of plug tray feeder lines in large-scale building pig rearing can largely avoid the entry of foreign vehicles and personnel into the pig farm and prevent the spread of external diseases in the pig farms, thus improving the biosecurity of the pig farms.

In plug line conveying, the performance of the stopper, the corner wheel, and the material line often affect the stability of the system unit [31]. There are two main types of stoppers: steel buckle and wire rope, and the performance of chain discs connected by steel buckles is more stable than those connected by wire ropes [32]. When the stopper line system starts working, the feed enters the conveying pipe through the vibrating hopper and is accurately conveyed to each drop unit by the push of the stopper. As soon as the level sensor detects that the unit is fully fed, the system stops working and switches to the empty feed operation mode. At this time, the system will clean up the residual feed in the pipeline to prevent it from impacting regular operation.

The workflow of the plug line conveying is shown in Figure 1.

Plug line conveying is highly adaptable in building pig barn structures and can be flexibly configured and customized according to different site conditions and production requirements. However, it also has shortcomings, such as feed clogging, poor accuracy, and short equipment life. To solve these problems, scholars at home and abroad have conducted several studies. To solve the problem of feed blockage during conveying, Gan Ling et al. proposed an optimized pipeline cornering device and compared it with the traditional cornering device, and the residual feed filling rate of the cornering wheel was reduced from 79.2% to 28% [33]. To solve the problem of low accuracy, Qiu Zheng et al. modeled the cylinder using Comsol software (latest v. 6.2) to improve the conveying accuracy of the system and reduce residual feed by applying appropriate external resistance under ideal conditions [34]. The economic efficiency of pig farms is closely related to feed clogging and feeding accuracy, but the equipment’s service life is also an important influencing factor. Aiming at addressing the problem of the short service life of the equipment, He Renzai et al. proposed a method of connecting the pulling rope with the stopper to prevent it from breaking due to the maximum stress and deformation at the connection between the wire rope and the stopper [35]. In addition, Lu Z. proposed a chain tension control party system based on a fuzzy adaptive algorithm. This system can automatically adjust the chain tension according to the load it is subjected to, reducing the wear of the chain under high load [36]. Additionally, the system is prone to chain clogging during work, which affects the worker’s assessment of the maximum performance parameters of the component. Therefore, Shprekher D M et al. developed a mathematical model of a scraper conveyor card, in the case of chain blockage, using the finite element method to assist experts in more accurately evaluating the component’s maximum performance value index [37].

Failures are difficult to detect during prolonged use of the equipment, which can lead to financial losses. Therefore, Zhao S et al. proposed a method for gear fault diagnosis under time-varying load conditions [38]. The gear failure characteristic frequencies are extracted by pre-processing the raw currents during operation to determine gear failures successfully under time-varying load conditions. This method has no sensor installation requirements, which can save costs. However, this method has only been tested for diagnosing gear faults, and its effectiveness for other faults is unknown. Some special parameter characteristics exist when the equipment fails.

Jiang S. et al. used a combination of experimental and simulation methods to analyze the characteristics of the sprocket wheel in the system in abnormal working conditions of the speed and contact force [39]. Bianchi M. presented a wear measurement system utilizing laser rangefinders to evaluate transmission equipment. The system predicts transmission chain failures in advance, facilitating timely repairs by maintenance personnel. However, the method in the literature is limited to detecting whether the drive chain has failed and is not necessarily applicable to diagnosing other faults [40]. Jun Mao et al. proposed a signal analysis method based on fuzzy clustering and wavelet transform to analyze and diagnose the vibration signals of the transmission system [41]. The method provides a new way to accurately determine the equipment’s operating status and fault types, facilitating maintenance and repair and extending its service life.

These studies provide guidance and direction for plug line conveying in three areas: feed clogging, low conveying accuracy, and short equipment life. To address the problems of feed clogging and short equipment life, the first thing to consider is to optimize the structure of the equipment. Secondly, timely detection of equipment failure is also the key to ensuring the plug line’s regular operation and preventing feed clogging. The problem of conveying accuracy can be adjusted from the parameters of the equipment itself, but the external resistance can also be studied to improve the accuracy of feed conveying.

2.3. Screw Conveying

Screw conveying refers to the fact that when the screw rotates, the feed particles move forward with the rotation of the screw [42]. Screw conveying is suitable for simple linear feed transportation but is less adaptable when faced with complex spatial layouts. Screw conveying has the advantages of strong conveying capacity, simple structure, easy installation, and affordable price compared to scraper-type pipe conveying [43]. Therefore, screw conveying has been widely used in large-scale pig farms, but the operating characteristics of screw conveying are susceptible to regional limitations, and its conveying performance is affected by several key parameters.

Screw conveyors have many disadvantages when operating in complex spatial layouts [44], such as the low flexibility of conveying materials in a complex workspace and the influence of the main parameters of the screw conveyor on the conveying performance. In response to the problem of flexibility of materials conveyed by screw conveyors, Chakarborthy et al. proposed a semi-flexible screw conveyor design incorporating gimbals, which can be connected at an angle according to the actual situation, adapting to the requirements of a complex spatial operating terrain [45]. Bulgakov et al. not only proposed the design of a flexible spiral-section working body operating in complex spaces to improve the adaptability of screw conveyors in complex spaces but also found that the dynamic loads of the system can exceed the critical torque of the motor by 50–70% during fast braking by constructing a mathematical model of the dynamic system [46]. In addition, in the study of the main parameters affecting the performance of the screw conveyor. Zhu et al. investigated the influence law of pitch on the performance of screw conveyors using the discrete element method [47]. The conveyor exhibited the highest efficiency when the spiral blade-to-pitch ratio was 0.67. However, as the ratio increased between 0.67 and 0.9, the conveyor efficiency decreased sharply and was ultimately lowest at a ratio of 0.9. Unlike the study of pitch on conveyor performance, Wang et al. used a discrete element method to numerically simulate and analyze the flow behavior of granular materials in a screw conveyor [48]. It was found that when the fill rate of the feed in the screw conveyor was 20%, and the screw speed was between 300 rpm and 1000 rpm, the conveying performance of the conveyor was directly proportional to the screw speed.

Additionally, the screw speed has a more significant effect on the feed mass flow rate compared to the fill rate. Sun et al. used the discrete element method (DEM) to simulate conveyors with different axial inclination angles and used the angle of repose as an evaluation index [49]. It was found that the axial tilt angle of the blades had a significant effect on the mass flow rate when the rotational speed was in the range of 30 rpm to 60 rpm, and the mass flow rate increased linearly as the axial tilt angle increased from −15° to 20°. Unlike Sun L et al., who used the angle of repose as the evaluation index, Wu et al. analyzed the average wear depth and maximum wear depth as the evaluation index and used the EDEM discrete element method and control variable method to analyze the effect of the screw diameter, rotational speed, fill rate, and pitch on the wear of the screw conveyor [50]. The design of optimal operating parameters in the case of ensuring the minimum wear performance of the screw conveyor provides a specific reference value for the design of the screw conveyor. Most of the above scholars have examined the effect of each parameter on the performance of the screw conveyor utilizing the discrete element method. However, the disadvantage of using the discrete element method to simulate the relationship between each parameter and the flow rate is that it requires much experimental data and time. Therefore, Kalay E. proposed a method for predicting mass flow rate based on DOE (Design of Experiments) optimization of ANN parameters. In addition, the method predicts the mass flow rate based on relevant parameters such as screw speed, particle diameter, and pipe diameter. It provides an effective approach to investigate the influence of the key parameters on the screw conveyor’s performance [51].

The post-failure detection and repair of screw conveyors are significantly crucial for the high productivity of an organization. Timely detection of faults in the device is conducive to improving the stability of equipment operation and ensuring its regular productivity. Therefore, Li Y. proposed a deep-learning-based method for predicting clogging phenomena in screw conveyors. The method detects and predicts the clogging events and their severity through current and vibration signals [52]. Gao et al. designed a laser-scanning-based online detection system for the welding quality of spiral conveyors [53]. The system’s single detection time is 1 min with a time consumption 15~20 min less than the traditional device detection methods.

The online detection system workflow is shown in Figure 2.

Advanced spiral conveying pipeline fault detection technology offers multiple advantages. It can not only find faults in the pipeline in time to avoid production stops and losses caused by faults but also be able to find all kinds of safety hazards in the pipeline in time to improve the safety of the whole delivery system and realize real-time monitoring of the pipeline status. Current fault detection techniques have come a long way, but further improvements are still needed. The future development trend is not only to detect faults faster and more efficiently but also to accurately locate the fault, diagnose the fault reason, and provide the corresponding repair program so that the staff can carry out timely repairs and prevent a similar error occurring the next time, to improve the accuracy and reliability of fault detection technology.

2.4. Rail Conveying

The rail conveying in multi-storey building pig rearing mainly refers to the erection of rail tracks above the pig pens in the piggery, where the transportation device loads and unloads the feed from the feed box into the troughs in each pig pen. Currently, the pipeline feed delivery method is used in most domestic multi-storey buildings for pig raising. Although pipeline conveying is highly efficient and easy to operate, its bends are easy to wear and tear, the working time is shorter, and it is prone to pipeline clogging and other issues [54]. In contrast, rail conveying not only overcomes the problem of bend wear but also minimizes feed residue, effectively improving on the shortcomings of pipeline conveying. Rail conveying is also highly flexible and can be adapted to the needs of complex spaces through flexible track layouts.

Ye et al. designed a rail feeder trolley for use in a two-row, enclosed pig house [55]. Double-track rails are erected above the pig pens, and the feeding bins are near the rails.

The workflow of the rail feeder trolley system is shown in Figure 3.

The automatic control system utilizes switches on the front of the trolley, bumps on the rails, and travel switches at the end of the rails to ensure that the feeder trolley can convey the feed from the refill bins to the troughs in each pig pen. However, two layers of material boxes are inside the feeder trolley, which is used for storing and pouring feed. The structure is complicated, and the outlet of the trolley is on the small side, which causes arching of the feed in the box and inconsistency in feed delivery. To address this problem, Pang et al. designed an automatic feed rail conveyor system [56]. The system consists of a weighing platform, a feeder trolley, a horizontal track, an elevated track, and a control system, which showed that the average error between the theoretical value of the dropped material and the actual amount of the dropped material is less than 1%. However, the speed of the rail conveyor for feed is slow, and the cruising speed of the rail conveyor in this paper is 0.334–0.569 m/s. Although the drop accuracy has improved greatly, its cruising speed is low. Aimed at this problem, Zhao et al. designed a track feeder trolley to control the whole process automatically using magnetic spikes and magnetron switches to start and stop automatically [57]. Its actual amount of falling feed and the average deviation from the theoretical value are the same as that of the rail feeder trolley designed by Pang, but its cruising speed is 1.5 m/s, much faster than the cruising speed of the latter. In addition, some research has been conducted on tracked pig barn feed haulers [58]. However, tracked pig barn feed haulers are relatively more demanding in terms of operating conditions, requiring the consideration of factors such as road conditions, obstacles, and the location of troughs in the barn, whereas rail conveying does not require too many of these factors.

The results of recent research on pipeline and rail conveying are illustrated in

Table 1. Comparison of feed delivery methods.

Conveying Methods		Advantages	Drawbacks	References
pipeline conveying	Pneumatic conveying	Strong conveying capacity No transmission parts, reducing the possibility of maintenance High layout flexibility	High energy consumption Higher feed crushing rate	[14,15,16,17,18,19,20,21,22,23,24,25,26,27,28]
	scraper pipeline conveying	Smaller feed crushing Flexible space layout	Weak transportation capacity Conveying viscous or wet feed, prone to clogging	[33,34,35,36,37,38,39,40,41]
	screw conveying	Affordable price Simple structure and easy installation	Poor layout flexibility Prone to feed clogging	[45,46,47,48,49,50,51,52,53]
Rail conveying		Flexible layout in the pig house Avoiding feed clogging Avoiding bend erosion	Slower tracking and less efficient transportation Low conveying efficiency Higher transportation costs for installing rails	[55,58]

, and several types of pipelines and rail conveying are compared separately.

3. Automated Feeding Systems

At present, most pig farms use pneumatic conveying, scraper pipeline conveying, screw conveying, and other methods of feed conveying. Large-scale pig raising has undergone the mechanization, informatization, and intelligence stages. In the mechanization stage, the system is an entirely isolated island that perceives no external information and can only perform standardized operations. Most pig-raising enterprises in China have entered the informatization stage but still rely on human experience, resulting in feed waste. A few companies have entered the intelligent stage, replacing traditional farming techniques and management experience with precise farming management systems [59].

As incomes and populations rise in developing countries, global demand for pork is gradually increasing [60]. The traditional manual feeding and single mechanized feeding methods cannot fully consider the individual differences and nutritional requirements of different pigs, resulting in the inability to achieve accurate feeding, which affects production efficiency [61]. In the current era of information technology, pig farm automated feeding systems can collect data on different growth periods and feeding conditions of pigs for personalized feeding management, improving feed utilization and promoting the growth and development of pigs [62]. By combining smart technology with pig farming, it is possible to monitor the health of pigs in real time, implement personalized feeding management, reduce the number of times staff frequently enter and leave the barn, and reduce the risk of invasion by external pathogens, thus improving the biosecurity of pig farms. The main feeding systems in the world today are the “Gestalt” feeding system, the Velos feeding system, the fattening pig partitioned feeding system, and the intelligent liquid feed feeding system.

3.1. “Gestalt” Feeding System

The “Gestalt” feeding system designed by JYGA Canada is the world’s first computerized sow feeding system, whose main components include software, a wireless receiver, and a feeder [63]. This feeding system is specially designed for lactating sows and automatically collects and analyzes the feeding data of lactating sows, then gives recommendations to the manager to meet the nutritional needs of lactating sows at different stages. This personalized management approach enables real-time monitoring of the health of lactating sows through feeding data [64]. The application of the “Gestalt” feeding system in the management of lactating sows helps to control the thickness of backfat and improve the health of sows during lactation [65,66]. However, there is still some error between the actual feeding quantity and the theoretical value.

The “Gestalt” feeding system is shown in Figure 4.

To overcome the problem of feeding accuracy, Xiong et al. designed an accurate automatic feeding controller for lactating sows [67]. The results produced by this device are very close to the feeding pattern of lactating sows and superior to manual feeding. Since the conventional discharging device will inevitably produce feed residue and other problems resulting in waste, Xiong et al. redesigned the discharging device and proposed a new low-cost precision discharging control system for lactating sows [68]. The system differs from other downfeed mechanisms using horizontal screw conveyors in that it primarily uses an electric actuator-based control mechanism that works in concert with an embedded system. Compared to the previously designed automatic feeding controller, this system improves the feeding effect and reduces the amount of feed residue in the unloading device. The feed intake data of lactating sows are one of the objective parameters for assessing their health status and an essential basis for predicting their next feed intake. Gauthier et al. proposed a mechanism model that combines an online prediction program with an offline learning program [69]. Using the K-Shape algorithm for clustering and offline learning on the existing feed intake data of lactating sows, the trajectory curves of their feed intake can be obtained, and in turn, the feed intake of sows on the following day can be predicted. This approach automatically helps the equipment update the daily feed intake for more personalized management and can be easily embedded into the Gestalt feeding system to provide more intelligent and precise feeding control for aquaculture. As this method only considered feeding data and did not give much consideration to the role of other factors in accurate feeding decisions, Gaillard C. et al. proposed a multifactorial correlated precision feeding system that assesses the actual nutritional needs of sows more accurately and gives staff the most realistic feedback [70]. Gaillard C. and Durand M. explored the influence of behavioral and environmental factors on the nutritional requirements of sows, providing a theoretical basis for developing more effective decision support systems in the future [71,72].

Research on the “Gestalt” feeding system will help to increase feed utilization, improve the nutritional status of lactating sows, and effectively avoid under- or over-supply of feed. Although the current “Gestalt” feeding system is mainly used in single-pen lactating sow housing, which restricts the free movement of the sows, it facilitates the intelligent management of the sows and helps prevent the spread of external diseases. Future research on Gestalt feeding systems could favor the integration of nutritional models with decision support systems (DSS), integrating and weighing data from different sensors in order to provide customized management solutions for precision feeding.

3.2. Velos Feeding System

The “Gestalt” feeding system developed by Neadap in the Netherlands for managing lactating sows is mainly based on restriction pens. However, lactating sows kept in restriction pens will be in a state of subhealth for a long time due to the lack of exercise in a small space and will be prone to suffer from reproductive disorders, which is not conducive to the physical health of the lactating sows [73]. The results of the study showed that the Velos feeding system had more significant advantages over the traditional restriction pen feeding system in terms of lowering the duration of labor and reducing the number of weak and stillborn fetuses [74]. Dutch research shows that the Velos feeding system can free lactating sows from restriction pens, freeing them from their herd nature and allowing them to move freely to improve sow health and reproductive performance [75]. The Velos feeding system involves installing chips in the ear tags of group-rearing sows. The chips contains the biological file information of the pig. When the sows want to feed close to the feeder, the equipment uses radio-frequency technology to identify the sow’s ear tags to determine whether the sow is in the estrus stage and then develop a feed delivery program to achieve personalized feeding [76]. The workflow diagram of the Velos feeding system is shown in Figure 5. The management of group-fed sows can satisfy the behavioral expression needs of sows, alleviate their stress response to a certain extent, and improve their reproductive performance.

In the Velos feeding system, the electronic ear tag is not only the identification of the pig but also the basis for the realization of intelligent group feeding of sows [77]. At present, low-frequency radio-frequency identification technology (LF-RFID) has been widely used in the pig industry and is recognized by most people as a more mature pig identification and management of standard technology [78,79]. Identifying electronic ear tags on sows through low-frequency radio-frequency facilitates personalized management to a certain extent. However, the read range and data transmission efficiency of LF-RFID are relatively low [80,81]. In order to make the Velos feeding system widely used in multi-storey building pig rearing, it is necessary to solve the problems of its small reading range of electronic ear tags, low efficiency of data transmission, high cost, and single function. Aiming at addressing the problems of small reading range and inefficient data transmission, Suriyasomboon et al. conducted a comparative study on the application of low-frequency (LF) and high-frequency (HF) radio frequency identification (RFID) ear tags in pig herds [82]. It is found that the difference between the two in terms of readability and retention rate is not significant, but the high-frequency RFID ear tag performs better in terms of reading distance and speed; its reading distance is farther, and its reading speed is faster, which have certain advantages compared with the low-frequency RFID ear tag. Considering that most of China uses Afikim’s equipment for electronic ear tagging, its cost is relatively high. Therefore, Jiang et al. designed a new type of electronic ear tag in which the UHF electronic ear tag chip, pedometer, and ZigBee communication module are integrated [83]. This design’s cost is significantly lower than that of purchasing Afikim equipment, thus reducing dependence on foreign technology. Electronic ear tags, which are widely used on the market, are mainly used to store individual pig information so that they can be accurately identified at feeding time, thus personalizing the day’s feed intake plan for each pig. Wang et al. expanded the role of electronic ear tags by combining the identification technology of electronic ear tags with GPS technology to realize real-time monitoring and management in the breeding process [84]. On this basis, Zhang et al. built a luminescent electronic ear tag test platform using two typical UHF electronic ear tags on the market at present and sent luminescent commands to find the location of the target sows to facilitate the localization of abnormal pigs [85]. Electronic ear tags are useful not only for locating abnormal pigs but also for assessing whether a pig’s daily feed intake is abnormal. Bruijin et al. proposed an early warning model for correlated variables based on RFID enrollment and Kalman filtering [86]. Through the use of Kalman filtering technology to predict the amount of food and water pigs eat and drink each day and its comparison with the actual eating and drinking behavior (FDB) of pigs, if the data of the pig’s electronic ear tag of the FDB deviate from the expected level, it will automatically generate alarms to send alerts to the staff. Although Bruijin’s RFID study focused on fattening pigs, considering the susceptibility of lactating sows to disease and, thus, their differing feed intake behavior (FDB values), we can try to apply the model to lactating sows for timely monitoring of their feeding status. Vargovic L. presented an electronic feeding system suitable for monitoring group-housed sows. The system monitors potential health problems in sows by fitting curves to the data and comparing them with the sows’ daily data [87].

Although the Velos feeding system has been developed to a certain extent, future research will mainly focus on the integration of RFID, GPS, and Kalman filtering technologies to achieve the functions of real-time tracking of the specific location of the pig, formulating a personalized feed feeding plan based on the information of the electronic ear tag, and instantly monitoring the pig’s feeding and drinking behaviors to provide timely reminders for the staff. This facilitates the timely prevention of epidemics and interrupts the spread of epidemics, thus protecting the herd. The Velos feeding system allows for group feeding of gestating and lactating sows, but some feel that group feeding increases physical contact between sows in the barn, which increases their exposure to feces and urine, leading to disease and reduced immunity. In response to this question, studies by Von et al., McGlone et al., and Broom et al. found that both individually housed sows and sows housed in groups produced the same antibodies- and neutrophils-to-lymphocyte ratios and that this did not have an effect on their immunity [88,89,90]. However, group housing is suitable for releasing the herd nature of sows to increase their exercise and improve their health. Therefore, increasing the research on radio-frequency technology to be widely used in the pig breeding industry to promote intelligent and informalized farming is necessary.

Table 2. Listing of research results on sow feeding systems.

Feeding Systems	System Design	Effect	References
“Gestalt” feeding system	An Intelligent feeding system for lactating sows combining feed Intake modeling with precise rotation of wiper motor technology	20.5% higher than manual daily feeding	[67]
	A new embedded precision feeding control system for lactating sows	22.4% higher than manual daily feeding	[68]
	A time series forecasting program based on big data, K-Shape clustering approach	Predicting the next feed intake based on the current day’s feed intake	[69]
	A multi-parameters correlated precision feeding system	Predicting next feed intake based on environmental and behavioral factors	[70]
Velos feeding system	Comparison of the performance of low and high frequency electronic ear tags using chi-square analysis	High-frequency electronic ear tag with better performance	[82]
	An electronic ear tag comprising UHF RFID chip, pedometer and ZigBee communication module	Reduced equipment costs and dependence on foreign technology	[83]
	An online monitoring method combining electronic ear tags and GPS technology	Easy to keep track of the general location of the pig	[84]
	A luminous electronic ear tag test platform	Easy targeting of pigs using luminescent commands	[85]
	An early warning model for correlated variables based on RFID enrollment and kalman filtering	Compare predicted and actual FDB values of pigs to determine if pigs are abnormal	[86]
	An electronic feeding system suitable for monitoring group-housed sows	Simulating feeding curves to monitor potential health problems in pigs	[87]

lists recent research results on sow feeding systems, providing a reference direction for later researchers.

3.3. Fattening Pig Partitioned Feeding System

The fattening pig group feeding system designed by Raybolt is a method of grouping fattening pigs in the same batch according to their weight size. In this way, it is possible to individualize the feeding management of pigs with different body weights. This helps to optimize feeding [91]. The fattening pig partitioned feeding system works as shown in Figure 6. When the pig enters the pen splitter, the device will automatically recognize the electronic ear tag on the pig, thus assigning it to different areas for feeding. In the Japanese pig breeding model, fattening pigs with different growth and development conditions are also separated into groups to ensure the balanced growth of the whole group [92]. It has been reported that the survival rate of fattening pigs raised in large pens is at least 2% higher than that of pigs raised in small pens and that it reduces the stress caused by fattening pigs being herded at the time of sale [93]. However, there are some problems with the fattening pig partitioned feeding system.

To address the problem of rapid separation of large-scale group rearing of pigs, Huang et al. proposed a separation device for group rearing of fattening pigs [94]. The dividing device includes several passages, and the two side walls of the passages extend from the exit section of the passages to the outermost two side walls of the passages, thus realizing the rapid separation of large groups of pigs. However, some objective conditions sometimes affect it, resulting in the impossibility of portioning, so Liu et al. have designed an automatic portioning system and feeding device for fattening pigs [95]. By dividing the feeding area in a large pig pen into two separate zones, the system can intelligently assign feeding pigs to different zones for independent feeding using wireless ear tag identifiers and electronic weighing devices. This design enables the effective implementation of group feeding even when traditional pen divisions are impossible. However, some partitioned devices may cause pigs to be trapped in feeding due to power outages. Aiming at addressing the problem of pigs being trapped, Liu et al. designed a livestock feeding partitioned feeding device and its electronically controlled door-locking mechanism [96]. The livestock feeding partitioned feeding device utilizes a lever mechanism that ensures that one of the polyaxial gates remains open, avoiding the possibility of fattening pigs becoming trapped in the passageway, whether due to a power outage or other malfunction. The performance of a partitioned feeding system for fattening pigs is influenced by several factors, of which the performance of the partitioned device is only one. Another equally important influence is the accurate estimation of the weight of the fattening pigs. Regarding fattening pig sorting, in addition to the use of electronic weighing devices, Reckels B. suggested that the ultrasonic detection of the ratio of backfat thickness to the diameter of the longissimus dorsi muscle in fattening pigs can be used for sorting [97]. Lengling A. suggested the application of infrared thermography as a non-invasive method to classify fattening pigs based on imaging characteristics, but relatively little research has been conducted [98].

To address the lack of accuracy in estimating body weight in a fattening pig partitioned feeding system, Brand et al. used image analysis techniques to measure the body dimensions and characteristics of pigs. Then, they combined these data with a pre-established body weight estimation model to infer the actual body weight of the pigs [99]. However, the error in estimating pig weights was about 5–6%, and there is still room for improvement. In order to minimize the effect of image overlap on body weight analysis, Kashiha et al. devised a pig weight method using top-view image processing [100]. The method utilizes an ellipse fitting algorithm to locate pigs in the image and avoids the effects of image overlap on weight prediction in the brand image analysis technique. Next, the area occupied by the pig in the ellipse is calculated, and modeling is used to predict the pig’s weight, with weight estimates averaging over 97% accuracy. In order to improve the average accuracy of weight prediction, Li designed a LabVIEW-based dynamic weighing and partitioned system [101]. Taking the dynamic weighing data of the system as the object, the BP neural network algorithm was selected to process the weighing data, and they were optimized using the particle swarm algorithm. The estimation error of the weight of each pig was within 1 kg, and the correct rate of partitioning was close to 100%.

Additionally, the mass of livestock is often closely related to their size and weight, so some researchers have begun to use point clouds for 3D model construction to estimate the size and weight of pigs. Kwon et al. proposed an iterative offset-based reconstruction method for pig-related point cloud mesh models [102]. By using the developed prototype system to process the existing pig point cloud data and reconstruct the network model to reduce the error between the model and the pig’s body size, the accuracy of the weight prediction in the experiment was as high as 97.57%. Pigs were evaluated for abnormalities based on the size and mass of the pig obtained from the reconstruction model.

The fattening pig partitioned feeding system has undergone some research and innovation regarding dividing devices and the pigs’ weight prediction algorithm, which, to some extent, facilitates the realizing of the automation mode. However, existing research on control systems for partitioned devices and weight prediction methods cannot determine abnormalities in feeding pigs. Therefore, the future research direction can consider trying to combine the partitioned device control system, 3D point cloud model construction technology, ultrasound technology, infrared thermography, and disease recognition technology to realize the goal of automatic color spray marking for pigs with abnormal behavior, heat, disease, and lost ear tags and automatically separating them into specific areas.

3.4. Liquid Feed Intelligent Feeding System

Liquid feeds are mainly used for weaned piglets and sows in the pig industry. Compared with solid feeds, liquid feeds help to improve the gastrointestinal condition of weaned piglets and sows, increase their daily feed intake, and relieve their stress [103,104]. Liquid feed was first used in Europe; so far, 30–60% of the feeding systems have adopted liquid feeding technology [105]. The intelligent liquid feed system comprises computer program control, batching and mixing, and conveying. The system is mainly computerized to accurately calculate the total amount of liquid required for each feeding circuit. Subsequently, according to the feeding formula, each feed ingredient tank proportionally conveys the ingredients to the mixing tank for mixing, and the mixed liquid feed is accurately conveyed to each discharging tank through the feeding pumps [106]. Its workflow is shown in Figure 7. When piglets are first weaned, many factors may adversely affect the intestinal tract, decreasing brush border enzyme activity and absorption capacity, indicating the piglet’s digestive capacity and decreasing daily feed intake [107,108]. Although liquid feed is more advantageous than traditional dry feed in promoting pig growth and improving feed conversion ratio, it has some problems, such as extended response time, significant errors in feeding accuracy, and the deterioration of residual feed [109].

To address the problem of a long response time of liquid feed intelligent feeding systems, Olaniyi et al. investigated the performance evaluation of a mobile smart poultry liquid feed distribution system by using a bi-directional controller approach with a genetically tuned PI controller and an internal mode controller (IMC) [110]. The application of a genetically tuned PI controller was found to impact system performance positively. Yilmaz M designed a Pl+ feedforward controller based on a genetic algorithm to maximize the system control sequence’s performance to improve the system’s stability and responsiveness [111].

Aiming at addressing the problem of large errors of feeding accuracy, Zhang et al. designed an automatic fine feeding system for piglets, taking STC89C52RC chip as the core and combining it with a liquid-level sensor, photoelectric sensor, and other components to realize the real-time monitoring of the state of the feed box and accurate feed delivery, and the error between the actual amount of feed and the theoretical value was within 5% [112]. Although the error has been reduced to 5%, there is still potential for further improvement. Li et al. designed a liquid feeding system based on STM32 [113]. The system uses RS-485 communication to coordinate the hardware platforms effectively, combining arithmetic mean filtering and least squares to construct a linear model. The feeding accuracy error is controlled at less than 1%, and it can send out an early warning signal when there is a malfunction, which helps the staff perform maintenance on time. The accuracy of this liquid feeding system has been significantly improved, reducing the error rate to less than 1% compared to the system designed by Zhang. In the application of large-scale feeding, if the cost of each ton of feed is 2307.7 dollars, then the system can be used to ensure that the feeding will not be affected under the premise of cost savings of 92.3 dollars per ton. This cost reduction effect not only improves the economic efficiency of the enterprise but also enhances its market competitiveness.

Although the liquid feeding system researched by Li has a feeding precision error of less than 1%, its text does not elaborate on the disinfection and cleaning process of the pipeline after the system completes the feeding task, which can easily lead to problems such as mold and germs if there are feed residues. To address this problem, Deng et al. proposed a liquid feed intelligent feeding system based on a programmable logic controller [114]. By comparing the feeding accuracy of the feeding system with domestic pumps and the feeding system with foreign pumps, it was found that the maximum error of feeding accuracy with domestic pumps was within 1.6%, while the maximum error of feeding accuracy with foreign pumps was within 0.7%. In addition, the system automatically disinfects and cleans the mixing tanks and conveying pipes after each operation to prevent residual feed from becoming moldy and thus affecting the next feeding. The above liquid feeds are conveyed by a pipeline, which inevitably leaves a small amount of liquid material in the pipeline, creating problems such as bacterial growth and contamination. To address this problem, Daqing et al. investigated an intelligent pig liquid feeder trolley with Zigbee communication and networked operation technology to automate loading, transportation, and feeding [57]. The feeder trolley realizes automatic positioning and precise parking through the magnetic control switch and uses the screw pump to realize the precise feeding of feed, and the actual feeding error is within 2.5%.

Although the liquid feed intelligent feeding system has been developed, it still needs increased investment in technology and the research and development of a more efficient, low-cost feeding system, to ensure small feeding precision error and high stability while shortening the response time. In addition, it is possible to study the early warning support system of the liquid feed intelligent feeding system to prevent the system from malfunctioning problems that may cause losses to the enterprise. Additionally, compared with the high-cost pipeline feeding system for liquid feed, the intelligent feeder trolley designed by Daqing et al. can effectively avoid feed residue and clogging problems compared with traditional pipeline conveying. However, its conveying efficiency and volume are low compared with liquid feed conveying systems. Therefore, there is a need to innovate in technology and develop intelligent feeder trolleys with higher efficiency and better stability.

It has been found that when the feed is mixed with water, the proliferation of natural lactic acid bacteria and yeasts in it produces lactic acid, which lowers the mixture’s pH, and lactic acid has a strong bactericidal effect on Enterobacteriaceae under weakly acidic conditions [115,116,117]. Lactobacillus content of more than 100 mm in the feed facilitates the intestinal digestion of piglets and reduces the incidence of salmonella [118]. Therefore, liquid fermentation feeding has gradually come to people’s attention, but the liquid fermentation feeding system has problems such as difficulty screening beneficial bacterial species, imperfect fermentation technology, and difficulty proportioning.

In order to study whether fermented grains are beneficial to the growth of piglets, Li et al. fed weaned piglets with fermented grain liquid and dry feed of the same formula, respectively, and the pigs’ average daily gained weight in the fermented liquid feeding group increased by 13.5%, which proved that fermented grain liquid feed could promote the growth of weaned piglets [103]. Many microorganisms are present during fermentation, including some strains that may have harmful effects. Therefore, it becomes crucial to screen for strains that are beneficial to the intestinal tract. Aiming at reducing the difficulty of screening beneficial bacterial species, Wang et al. proposed a set of automated feeding equipment for liquid fermentation by the cyclic discharging method [119]. With this equipment, it is possible to screen the species suitable for pig farms and to realize the precise dosing of the equipment and its delivery through the pipeline. However, the intelligent feeding system on the market has a large error in controlling the ratio of liquid fermented feed, which easily leads to low quality of the finished product. Aiming at addressing the complex problem of proportioning fermented feed, Luo et al. developed a novel liquid feed fermentation device [120]. The device is designed with multiple additive tubes to monitor the weight of the feed in real time by installing a weigher at the bottom of the mixing tank. At the same time, the equipment also adopts a double-tank fermentation system, which significantly reduces the time of the production process. Luo and others also equipped precise temperature thermometers to monitor and regulate the temperature to increase productivity and ensure the fermentation process runs smoothly. Although fermented liquid feed, as an emerging feeding farming technology, has many advantages regarding the growth and gut health of weaned piglets, it is susceptible to external environmental influences. In winter, liquid feed may lead to high humidity in the barn, increasing the risk of pigs catching cold.

In contrast, high temperatures in the summer can lead to problems with residual feed spoilage and bacterial growth. In addition, the low temperatures and sterilization of pipes in winter are not suitable for the growth of lactic acid bacteria [121,122]. The fermented liquid feed has a good development prospect, and researchers may consider focusing their research on simplifying and optimizing the screening process of beneficial bacterial species, the ratio of fermented feed, and the fermentation technology of liquid feed.

Table 3. Development and comparison of feeding systems.

Feeding Systems	System Design	Effect	References
Fattening pig partitioned feeding system	A kind of group rearing fattening pig partitioned device	Realization of rapid separation of large-scale group pigs	[94]
	An automatic portioning system and portioning feeding device for fattening pigs	Personalized group feeding when partitioning is not possible	[95]
	A livestock feeding partitioned feeding device	Avoiding pig entrapment in case of power failure	[96]
	A method for classifying fattening pigs based on ultrasonic detection	Classification is achieved by the ratio of backfat to the diameter of the longissimus dorsi muscle.	[97]
	A classification method based on infrared thermal imaging technology	Non-invasive, non-contact classification, fitting fattening pig temperature profiles for monitoring	[98]
	An image analysis system using a hybrid model	Accuracy of weight estimation is about 95%	[99]
	A pig weight method using top view image processing	Accuracy of weight estimation is about 97.5%	[100]
	A dynamic weighing and partitioned system based on LabVIEW and particle swarm algorithms	Weight estimation within 1 kg	[101]
	An iterative offset-based reconstruction method for pig-related point cloud mesh models	Accuracy of weight estimation is about 97.57%	[102]
Liquid feed intelligent feeding system	A bi-directional controller method using genetic algorithm to tune PI controller and internal mode controller	Positive impact on system transient response and AE controller evaluation aspects	[110]
	A genetic algorithm based Pl+ feedforward controller	Improved system responsiveness and stability	[111]
	An automatic fine feeding system for piglets	The accuracy of the feed is about 95%	[112]
	An accuracy control algorithm combining arithmetic mean filtering and least squares method	The accuracy of the feed is over 99%	[113]
	A liquid feed intelligent feeding system based on programmable logic controller	The accuracy of the feed is over 98.4%	[114]
	An intelligent pig liquid feeder trolley	The accuracy of the feed is about 97.5%	[57]
	Designing experiments to compare the effects of liquid and dry feed on piglet growth	Average pigs daily gained weight in the fermented grain liquid group was increased by 13.5%	[103]
	A set of automated feeding equipment for liquid fermentation by cyclic discharging method	Average daily weight gain reached 0.907 kg and morbidity was reduced to 1.5 percent	[119]
	A kind of double-tank liquid feed fermentation device	Effectively alleviates problems with feed rationing and temperature detection	[120]

lists the recent research progress of fattening pig partitioned feeding systems and liquid feed intelligent feeding systems. The following conclusions can be drawn from the comparison:

• Although the accuracy of fattening pig partitioned feeding systems has been relatively high, it incorporates 3D point cloud modeling to predict the size of the pig to determine whether it is normal or not. However, no special studies have been conducted on abnormal pigs, such as estrus, illness, and loss of ear tags, and on the use of ultrasound technology and infrared thermography to classify fattening pigs. Future research should incorporate infrared thermography to achieve a fast and efficient penning system capable of automatically identifying and marking abnormal pigs and separating them automatically into specific areas. This will enable the tracking and handling of abnormal pigs to ensure that they receive appropriate attention and management.
• Despite the relatively low downstream error of liquid feed intelligent feeding systems, pipeline conveying systems are more costly. In contrast, the cost of the liquid feeder trolley is lower, but its transportation efficiency is not as high. Additionally, although the emerging fermented liquid feed combines the advantages of beneficial bacterial strains and liquid feed, which can significantly increase the daily feed intake and improve the body condition of pigs, there is relatively little screening and research on beneficial bacterial strains and the exploration of fermentation technology. Future research should focus on optimizing and simplifying the screening process for beneficial strains while ensuring the accuracy of the liquid feed ratios and reducing the effects of different seasonal temperatures on fermented liquid feeds. This can improve the efficiency and consistency of feed fermentation, further optimizing the quality and stability of liquid feed.

4. Challenges and Developments

Multi-storey building pig rearing is a significant attempt to reform China’s pig industry, which is still in the exploratory stage and faces many challenges and difficulties. They mainly include the following:

• Feed conveying machinery has problems such as frequent maintenance and poor layout flexibility, making it challenging to meet the feed transportation needs of building pig raising. Conveying machinery for feed is mainly divided into four types: pneumatic conveying, scraper pipeline conveying, screw conveying, and rail conveying. Among them, pneumatic conveying, scraper pipeline conveying, and screw conveying all belong to pipeline transportation and are favored for their efficient feed delivery. However, they all have drawbacks, such as feed residue and pipe wear problems. To overcome these problems, rail conveying was introduced and effectively improved upon the shortcomings of pipeline conveying. However, in general, rail conveying is only applicable to feed delivery at the same level, and its low transportation efficiency makes it difficult to meet the needs of large-scale feed delivery at the industrial level.
• Automated feeding systems mainly include the “Gestalt” feeding system, the Velos feeding system, the fattening pig partitioned feeding system, and the liquid feed intelligent feeding system, which, despite the high feeding accuracy, still have some problems. The “Gestalt” feeding system and the Velos feeding system have use limitations, are highly costly, and do not allow for good monitoring of the body condition of the sows while they are being fed. Although the fattening pig partitioned feeding system performs better in pen accuracy and is suitable for large barns, the system is currently unable to achieve automatic color spray marking and the special treatment of abnormal pigs, which limits the ability to track and deal with abnormal pigs. Although the liquid feed intelligent feeding system helps to improve the daily feed intake and weight gain of pigs compared with the solid feed feeding system, it has high equipment costs, inaccurate feed ratios, difficulties in screening beneficial bacterial strains, and the liquid fermented feed is susceptible to the effects of temperature in different seasons. These factors can cause instabilities in the quality of liquid feed.

In summary, in order to break through the limitations of the existing feed delivery systems, researchers should focus more on the following aspects:

• Explore new feed conveying machinery with many applications, good stability, and high erosion resistance. Study the working principles of pipeline transportation and rail transportation, combine the advantages of both, exploit the strengths and avoid the weaknesses, and develop new transportation machinery with a wide range, good stability, and high corrosion resistance to realize the least amount of pipeline residue in the transportation of feed as well as the least impact on the erosion of pipeline bends. For pipeline transportation, explore new structures for bending pipes to reduce feed residue. For rail transportation, try to explore the large-capacity loading silo to improve transportation efficiency. In addition, the use of current signals and vibration signals for equipment fault diagnosis is also an important exploration direction.
• Explore high-precision, low-cost, and widely adaptable solid feed feeding systems. High accuracy means higher feeding accuracy and less feed residue than existing discharge systems. Low cost means relative cost reduction while maintaining or improving the performance of existing automatic feeding systems. Existing automated feeding systems have limitations on what they can use for cost reasons, resulting in pigs needing to adapt to different feeding equipment at different stages, which can easily trigger stress reactions. Therefore, studying the solid feed feeding system with a wide range of adaptations and low cost is essential for liquid feed intelligent feeding systems for process simplicity, low cost, and accurate dosing. In addition, exploring a feed prediction system with multi-parameter correlation is likewise a key research direction to achieve personalized and precise feeding regimens. Currently, the liquid feed intelligent feeding system is not perfect and is high-cost, making it difficult to popularize in real life. However, liquid feed helps increase the daily intake of pigs and improve their gastrointestinal system, so future research on intelligent liquid feed feeding systems will be more likely to reduce costs and improve the process of liquid feeding systems.

In addition, pig feed conveying and transfer with separate power sources leads to system design complexity and high maintenance costs. Therefore, an attempt can be made to explore an integrated device that couples pig feed transportation and hog transfer together, sharing a common power system.

5. Conclusions

As manual feed transportation in large-scale pig farming increases farming costs and poses a potential risk to biosecurity, this paper summarizes the current state of the pig industry and identifies building pig farming as a future trend. This paper then describes several mechanically conveying pig feed methods and analyzes their limitations and problems. Finally, future trends in feed transportation are discussed. In addition, the potential of the current coupling of pig feed conveying and pig trans-shipment is also prospected to provide a reference and insight for researchers.

Different feeding systems are used for different sexes and stages of pigs, but their common shortcoming is a lack of precision and a low degree of personalized management. At present, pipeline conveying is widely used in pig farms, but it has problems such as feed residue and bend erosion. Among the several ways of pipeline conveying feed, pneumatic conveying has excellent conveying capacity compared with scraper pipeline conveying and screw conveying, but its energy consumption is higher. Although scraper pipeline conveying has low capability, it offers much layout flexibility. Screw conveying is the most cost-effective option, although it has relatively low conveying capacity and layout flexibility.

Railway conveying is a good solution to the shortcomings, but its conveying efficiency is low and only applies to the same level of operation. In addition, various feeding systems for large-scale pig rearing are presented, such as the “Gestalt” feeding system, the Velos feeding system, and the fattening pig partitioned feeding system.

It is crucial to study intelligent feed delivery to promote the automation and intelligent development of pig farms. This technology enables personalized and precise feeding, reduces frequent staff operations, reduces the risk of pigs contracting diseases, and improves biosecurity. Therefore, professionals should further research and develop related technologies to improve smart feed delivery reliability and efficiency to ensure healthy, safe, and efficient pig farming and achieve lower operating costs. Also, infrared thermography might help the penning system automatically identify and separate abnormal pigs.