Open Access is an initiative that aims to make scientific research freely available to all. To date our community has made over 100 million downloads. It’s based on principles of collaboration, unobstructed discovery, and, most importantly, scientific progression. As PhD students, we found it difficult to access the research we needed, so we decided to create a new Open Access publisher that levels the playing field for scientists across the world. How? By making research easy to access, and puts the academic needs of the researchers before the business interests of publishers.
We are a community of more than 103,000 authors and editors from 3,291 institutions spanning 160 countries, including Nobel Prize winners and some of the world’s most-cited researchers. Publishing on IntechOpen allows authors to earn citations and find new collaborators, meaning more people see your work not only from your own field of study, but from other related fields too.
Digital agriculture is increasingly being incorporated into all areas of agricultural production, but using different names. Names, including precision agriculture, digital horticulture, and precision forestry, are used within row crop, horticulture, and forestry systems, respectively. Within livestock production systems, digital agriculture is commonly called “precision livestock farming” or abbreviated PLF. The application of digital agriculture, or PLF, in swine production systems involves the digitization of all aspects of production. Examples of PLF within swine production systems include feeding, watering, ventilation, environmental control, disease detection, animal welfare, and many additional aspects of animal health management. A wide variety of sensors and algorithmic systems are used for real-time monitoring and control in pig production systems. Methods, including computer vision, sound, temperature, and movement sensing, are used to collect data that is in-turn analyzed by digital systems in order to make management decisions in real time. This chapter provides an overview of these applications and describes the current state of the science regarding the use of PLF via digital agriculture in swine production.
The University of Tennessee, Knoxville, Tennessee, United States
Robert Spajić
Faculty of Agrobiotechnical Science Osijek, Osijek, Croatia
*Address all correspondence to: rburns@utk.edu
1. Introduction
Modern pork production includes the use of digital agriculture techniques in all phases of production. The use of digital agriculture in pork and other livestock and poultry production is typically referred to as precision livestock farming or PLF. Precision livestock farming can be further subdivided into sensing, algorithm processing, and control functions. This chapter provides information on the current application of PLF in conjunction with swine feeding, watering, ventilation, environmental control, disease detection, animal health, and animal welfare systems.
Sensing is the collection of data regarding the production system components, (i.e., feed or watering system data), environmental parameters (i.e. temperature, humidity, or air quality measurements), or animal-based measures. Animal-based measures (ABM) can be defined as the responses from an animal, effects on an animal, and can be collected directly on the animal or indirectly from records [1]. Examples of animal-based measurements include direct measurements such as animal temperature, movement activity, and heart rate. Other direct measures, such as computer vision images and audio recordings, are also collected and may be used to determine animal body condition, Health, or welfare status.
Data may be collected using fixed, mobile, or wearable sensors. Examples of fixed sensors include typical flow or mass-based water and feed monitoring systems, temperature and relative humidity sensors, computer vision systems, microphones, ammonia, and carbon dioxide sensors. It should be noted that some fixed sensors, such as feed and water intake and computer vision systems, may be connected with individual animal ID systems in order to associate specific data to individual animals. Mobile sensors are simply fixed sensors that are attached to either robot or rail systems that can travel within an animal housing system. Wearable sensors are sensors that are attached or worn by an individual animal. In swine systems, wearable sensors are typically associated with ear tags or collar systems that incorporate activity sensors, temperature sensors, microphones, or other devices [2].
Analysis is the processing of the collected data via computer algorithms in order to make real-time decisions regarding the animal production process. Data analysis is often carried out via cloud computing such that farms do not have to possess on-site computers to process the data and in order to protect the intellectual property associated with the algorithms themselves. Currently, questions around data ownership, how data are used, and who has access to the data can be a contentious issue between farmers and PLF providers. Given that the data is often visual images that are stored in the cloud, and concerns about cyber security are also an issue [3]. Researchers have reported that continuing concerns over data ownership and security have resulted in a lack of trust by farmers that has in-turn slowed the adoption of PLF in commercial farm settings [4, 5].
Control encompasses automated actions to manage the animals or their environment based on the analysis of the sensing data. Control examples include the use of automated systems to operate ventilation or heat systems, or robotic systems to move or collect animals or remove bedding or manure. Not all PLF systems include an automated control process, but all do include sensing and analysis functions.
Commercial swine feeding may utilize either dry or wet feeding systems. In either system, it is important to quantify the mass of feed delivered to as well as actually consumed by the pigs. Variations between different production categories and phases will determine the type of the feeding system applied. Additionally, different feeding systems and techniques should be utilized in sow and nursery production units [6].
The use of PLF can provide significant benefits in both the sow/piglet and finishing pig production stages. Dry feeding systems are the most commonly used systems in sow/piglet units. Individual feeding for each sow is crucial in modern sow/piglet operations. Precision livestock farming systems provide producers with critical information on how much feed intake per day and per ration is consumed by the animals. Precise, automatic feeding for each production phase, and for each animal category, is required. There are three phases of the sow production cycle that are continuously repeated through the life of a sow, insemination, gestation, and farrowing. Producers should focus on an individual feeding approach in each stage of production, if possible, to optimize pig performance. Digital applications make this level of detailed feeding possible and manageable [7].
During the insemination phase, sows are typically housed in individual cubicles where each sow is held separately and fed through a period of 28–35 days. Feed is provided to the sow through individual mechanical or automatic feeders. Feed is offered to the sow through the feeder that has a feed box above each individual sow’s head. The planned volume of the feed will be delivered to the sow, usually three times or more per 24-hour period. Feed is provided to individual sows using a digital system that allocates feed to each sow by using the chamber and delivery tube as shown in Figure 1.
Figure 1.
Digitalized feeding units in insemination area (Picture Credit R. Spajic).
Normally feed rations are divided throughout the 12- to 16-hour daylight hours. Due to shorter days, the feeding process will include some night hours in the wintertime. Digital systems are used to monitor feed provision to each sow. As shown in Figure 1, each animal feed chamber has a “floating cap,” which will be set to a specific feed volume based on a body scoring during the sow’s first day in the insemination area. Based on the sow body condition score, each feeder unit will automatically deliver the amount of the daily feed required for the animal’s body condition score level that is based on data for each animal in the barn. The goal is that all the animals have a similar body score prior to their grouping in the gestation boxes/groups. The entire process is managed and monitored with the use of automated feeders controlled by a computer system.
Monitoring of feeding intervals and feeding volumes is determined by the sow parturition interval, animal genetics, seasonal weather conditions, etc. (as an example: feeding model in summertime where feed is divided into three rounds daily—40% of the day ration at 6 AM, 20% of the day ration at 11 AM, and 40% of the day rations at 5 PM). During the insemination phase of the sow production cycle, it is easy to control feed intake because the animals are usually confined either continuously or for several times a day. When the animals are confined, control of the amount and type of the feed provided is straightforward and can be monitored via mass or volume sensors. Automation of feed intake can be organized such that sensors that measure the outside environmental temperature will automatically determine the spread of the total ration’s volumes for the total day’s feed intake.
Gestation is the next phase of sow production, and the majority of modern swine operations keep the animals in a group box (this is currently mandatory in the EU). Typically, 10–40 animals are placed together, and they have an area where they can move, walk, and lay on the floor freely. Usually after a few days of hierarchy adjustments among the group of animals, each animal will recognize and determine the order of feed intake from automatic feeders. While there are farms where feeding of the gestating sows occurs directly on the floor, we will be focused here on the automatic feeder units and automatized feeding processes where digital systems are utilized. Figure 2 shows individual feeders for group gestation areas.
Figure 2.
Automated feeders in gestation area (Picture Credit R.T. Burns).
Mixing water with the feed in the trough is a common practice in the grouped animal sections. Figure 3 shows an example of gestation group feed through with individual animal feeding locations. Each animal approaches a feed trough and waits until the feed is delivered into the trough. The amount of the feed is delivered from the feeder is specific to the animal based on the recognition of the animals’ electronic responder identification. Responders are typically radio-frequency identification (RFID) units that are usually placed in the pig’s ear or on their leg.
Figure 3.
Individual pig feed trough in group housing (Picture Credit R. Spajić).
As mentioned above, in an automated feeding system, each animal is equipped with an individual responder. The reader will allow only one animal to enter the cubical where feed will be automatically delivered into the trough to feed a specific animal. A computer will monitor and control the frequency of the animal’s arrival at the feeding point on a daily basis. Each animal will be provided with the amount of feed suitable for its production status, age, body condition, etc. Some standard ranges of feeding intake volumes for production sows are presented in Table 1.
The farrowing production phase is the most sensitive production segment of the production cycle. Feeding models and feeding technology in this phase vary by genetics and age but are primarily defined by the number of piglets farrowed in parturition cycle for each individual sow. It is crucial that we have feeders that provide the exact amount of feed at the right time. Feeders are usually placed in the corner of the farrowing cubical as shown in Figure 4.
Figure 4.
Automated feeders in sow farrowing area (Picture Credit R.T. Burns).
Similar feed delivery technology is used during farrowing as was described for the gestation area. Some technical modifications of sow cubicles allow animals to move freely in the box, but the feeder should be still oriented and placed in the corner of the box for each individual sow.
The automated feeding systems described are in use at the sow/piglet farm units using dry feeding systems. When wet feeding systems are utilized, the major differences are related to preparation and delivery of feed. The dry components of the feed are mixed with the liquid substrates and water and then delivered to the animal as wet feed through the daily rations. Preparation of the feed is performed in the “mix kitchen,” where all the components are prepared, so the wet feed can be delivered by the pumping system to the animals. Computer systems are used to monitor and control feeding in order to reduce the manual steps in the whole process.
When digital feeding systems are used in finishing farm units, it is important that every step of the process is organized on a finisher group basis. Feeding processes in finishing operation units are usually split into two or three phases during the finishing process itself (after the weaning phase, mid-finishing phase, and final finishing phase). European finishing target live weights will vary from 110 to 115 kg, while finishing target weights in the United States will range from 125 to 140 kg of live weight. As previously mentioned, the feeding process in finishing barns is primarily organized on a group basis, where each group of pigs has their own ration delivered three to five times a day. The entire process is controlled by computer system, which provides hourly/daily status on feeding regimes for each animal group. The calculated amount of feed consumed on a daily basis, per each group of animals is the primary information used to monitor feed conversion in modern swine production systems. Based on the data provided by these digital systems, farmers can predict their business success for each production phase. There is also great potential to digitize the accurate delivery of feed additives and microelements through the animal feed in a precise manner to each individual animal.
Feeding and associated production data delivered in real time provides the information necessary to implement corrective interventions during each step of the production process. The key to this digital process is to minimize the number of parameters monitored needed to react in real time, so production interventions can be made as needed. Producers need to be cautious not to over-collect unneeded data that will not be used and are costly to both analyze and store.
Water is required in greater quantities than feed by pigs. Pigs typically require water-to-feed ratios of 2:1–3:1 during the nursery and grower-finish production stages pigs, and this ration declines as pigs grow [9]. Digital agriculture can be used to both control and monitor water provision in pig production systems per animal category/group daily. While it is typically recommended that water should be ad libitum provided to pigs, it is valuable to track water consumption by animals. The addition of water to feed is also used to closely manage the mass of feed consumed by sows in some instances. Commercial swine farm units utilize significant amounts of water. Water consumption should be monitored individually for each animal if possible.
Water is used for various purposes on swine farms, including animal drinking, cleaning, animal cooling, and for some air cleaning processes. Cleaning water significantly impacts the total water volume used on the farm. Usually, consumption of water is presented as a number of liters per kg of feed. It depends on the age of animal, stage of production, and climatic conditions [10].
Water use varies when comparing sow and finisher farms. During finishing, animals consume water necessary for their growth, while sows require water both for their own growth and for the growth of their piglets by producing milk for suckling. As an example, typical water consumption on Spanish pig farms categorized by pig category is shown in Table 2.
Range of water usage per different pig category in Spanish farms.
Includes all offspring of the sow until end of fattening period [10].
Monitoring of total farm water usage can be automated using digital systems. Significant water spillage or loss can be quickly identified with real-time water monitoring by installing sensors in different sections of production barns. Water level in waterers can be automatically checked through the use of sensors, and when the water level decreases an automatically operated valve membrane, as shown in Figure 5, will be opened and fill the waterer to the required level. This process is digitally controlled and monitored for each group of animals. A technical variation of this model is the mixing of feed and water together. The delivery of both the liquid and dry components should be monitored separately as it is more difficult to measure the slurry that is created after mixing the feed and water due to the mixture’s consistency.
Figure 5.
Electronic water delivery valve.
Monitoring of water intake by each individual animal is difficult to implement where animals are placed into groups. Equipping animals with individual identification and observing the individual animals drinking behavior and water is regularly done at a research level but is not yet commonly applied in production settings where the animals are maintained in groups.
Significant differences in water use on swine farms occur in the summertime compared to the wintertime. A large part of summertime water use can associate with cooling systems for swine farm operations (especially sow farm operation units with farrowing, and insemination areas) during hot weather. Precision water application using sprinkler systems is crucial for heat stress reduction during the farrowing and insemination phases. Overapplication of cooling water can cause high humidity rates in the barns, which can be also stressful for the pigs, as well as dilute stored manure. This is an additional reason that precision water application systems are preferred.
Different drinker configurations can have a significant impact on water usage. When drinkers are installed in less than optimal positions, or the number of drinkers is not adequate for the animal group size (regarding their phase of production, live weight, category, etc.) unnecessary water spillage can occur. To avoid this issue, it is very important to carefully design water provision systems and their installation for each segment of production system to ensure adequate water provision with minimum wastage. An example of nipple waterers inside drinking bowls that are rigidly fixed against a wall is shown in Figure 6. This approach minimizes water spillage.
Figure 6.
Properly installed wall-mounted water drinkers designed to reduce (Picture Credit R.T. Burns).
Air is exchanged in confined livestock production buildings to regulate the livestock production area temperature and humidity, as well as to maintain good air quality for animal production by removing carbon dioxide and ammonia. Pig production in controlled environment buildings began in the United States during the 1980s, and by the 1990s, the majority of pigs were produced in controlled environment buildings with mechanical ventilation systems [11].
There are several standard ventilation systems in use in controlled environment barns. Systems types include diffuse ventilation systems, direct airflow inlets, and preheated/precooled airflow inlets. The diffuse ventilation system is based on the inlet of the air above the roof panels in the barns where air is coming into the barns through the insulated ceiling with perforations to allow air entry. Air enters the barn through the small, perforated holes placed above the animal space. Direct airflow is engineered in such a way that air is drawn into the barn area via fans on the side of the barn walls. Air inlets are opened or closed based on the static pressure measurement inside the barn. Preheated or precooled air is sourced in this system via the pre-heating or pre-cooling of the incoming air in corridors or rooms on the barn’s sides where the air is later delivered to each barn production room separately. Many variations of the mentioned systems can be observed in commercial settings.
The initial application of digital agriculture was the use of computer systems to control ventilation management within swine production houses. These systems utilize temperature and relative humidity measurements to control the operation of ventilation fans in order to maintain the appropriate thermal neutral zone (TMZ) in the piggery that is neither too hot nor too cold for the animals. Current controlled environment pig production systems continue to use computer systems to manage the production environment within the house. Figure 7 shows a combined temperature and relative humidity sensor located inside a sow barn. The data from these sensors is used to control the barn ventilation system fans.
Figure 7.
Temperature and relative humidity sensor (Picture Credit R.T. Burns).
While temperature and relative humidity were traditionally the only parameters used to control barn ventilation, PLF systems are now in place that measure gaseous concentrations and ventilation fan flow rates as well. Figure 8 shows an exhaust fan configured with an airflow measurement system. Some PLF systems may also employ ammonia, carbon dioxide, and hydrogen sulfide sensors as well.
Figure 8.
Exhaust fan configured with airflow measurement device (Picture Credit R.T. Burns).
In addition to fixed sensors used to monitor environmental conditions inside swine barns, mobile sensors are also available. Mobile sensors may be mounted on rail, robotic systems, or handheld. Rail-mounted systems utilize a traveling unit mounted on a stationary rail above the pigs. The unit travels on the track or “rail” in a defined pattern around the barn. Handheld sensors or sensors worn by workers are also used within swine production systems to monitor environmental conditions within barns. Figure 9 shows a handheld thermal imaging camera being used to look for areas with a significantly different temperature than other areas of the animal housing area. Thermal cameras can be used to quickly detect either hot or cold spots within a barn caused by damaged or missing building insulation, defective heater thermostats, incorrectly located heaters, or poor ventilation uniformity. While handheld digital tools are not used for continuous monitoring and may lack high-level accuracy, they provide producers with a quick and lower-cost method to identify problem areas within a barn.
Figure 9.
Handheld thermal imaging camera (Picture Credit National hog Farmer - https://www.nationalhogfarmer.com/livestock-management/four-plf-technologies-that-will-help-manage-barns-better).
5. Disease detection, health management, and welfare monitoring
The outbreak of disease in swine production systems can have a tremendous negative impact on animal health and well-being, and in-turn on the profitability of the production enterprise. For example, the 2019 African Swine Fever (ASF) outbreak in China is estimated to have resulted in over $111 billion US dollars of economic loss in 2019 [12]. The rapid detection of disease in swine production systems can greatly reduce the negative impact of disease, as well as the economic loss induced by disease outbreaks. Because of the significant economic benefit associated with early disease detection, many PLF systems are currently available and marketed for this purpose. A recent literature review on PLF system use in swine production noted that accelerometers, load cells, computer vision, microphones, photoelectric sensors, pyrometers, and radio-frequency tracking sensors are being utilized for monitoring pig behavior, physical condition, and various health indicators in an effort to provide early disease detection [13].
There are currently multiple commercial PLF systems commercially available to pig producers that use either computer vision or microphones. Figure 10 shows deployment of a microphone that is used to collect data used to provide early detection of respiratory diseases in pigs. This system (marketed as SoundTalks by Boehringer Ingelheim) collects sound data with microphones and transfers the data to a computer algorithm in the cloud. The algorithm is used to isolate coughs and sounds typical of pigs with respiratory disease and notify an attending veterinarian once a threshold determined by the computer algorithm has been crossed.
Figure 10.
Microphones are used to collect data for early respiratory disease detection in pigs (Picture Credit: National hog Farmer / Boehringer Ingelheim https://www.nationalhogfarmer.com/farming-business-management/boehringer-ingelheim-unveils-new-respiratory-disease-detection-product).
This is an example of one of the many PLF systems that are currently commercially available to pig producers. Systems that provide early detection of respiratory disease using computer vision to monitor respiratory rate and heart rate have also been developed [14]. The goal of these systems is to inform either the veterinarian or farmer by the detection of the disease several days faster than it would have been recognized without the PLF systems.
Digital agriculture is used in a variety of ways other than early disease detection to assist producers in monitoring heard health and improve animal welfare. Indicators of health such as body condition score or animal posture can be monitored using computer vision systems. Animal feeding and drinking behavior, as well as feed and water intake, can also be monitored with digital systems.
Precision livestock farming systems have been recognized for their ability to provide real-time assessment of animal health and welfare status [15]. Proponents of PLF technologies suggest that real-time monitoring further contributes to the improvement of both pig health and welfare. Others believe that the adoption of PLF technologies could either directly or indirectly reduce pig welfare by causing direct or indirect injury to animals because farmers become over-reliant on PLF and lose their animal husbandry skills [16]. With the recognition that optimized animal productivity does not always equate to improved animal welfare, it is not surprising that there is some disagreement on the role of PLF in swine welfare.
The positive impact of digital agriculture and PLF technology in pig production systems has been well documented. Most, but not all, agree that digital agriculture and PLF will provide similar positive benefits to improve pig welfare as well.
The application of PLF in swine production is increasing in commercial pork production systems worldwide. Computers have been used to automatically adjust ventilation settings based on temperature set points for many years. Systems utilizing computer vision and sound to better monitor animal health and welfare are now being used in swine production systems on a regular basis. Wearable sensors, which have been applied far more regularly in the cattle industry, are beginning to be used on a more frequent basis in the swine industry as sensor cost decreases. Early disease detection via PLF systems is expected to provide the greatest positive return on economic investment. The monitoring of production conditions, as well as pig behavior and body condition, has been demonstrated to improve health monitoring in pigs. There are conflicting views as to the role that PLF has in improving pig welfare. A review of current publications on swine PLF indicates that the majority of scientists publishing on this topic believe that PLF will facilitate improved pig welfare going forward, but there are also scientists that warn of the potential negative effects on pig welfare through the adoption of digital agriculture and PLF in animal production systems.
References
1.Brscic M, Contiero B, Magrin L, Riuzzi G, Gottardo F. The use of the general animal-based measures codified terms in the scientific literature on farm animal welfare. Frontiers in Veterinary Science. 2021;8:634498. DOI: 10.3389/fvets.2021.634498
2.Pandey S, Kalwa U, Kong T, Guo B, Gauger PC, Peters DJ, et al. Behavioral monitoring tool for pig farmers: Ear tag sensors, machine intelligence, and technology adoption roadmap. Animals (Basel). 2021;11(9):2665. DOI: 10.3390/ani11092665
3.Morrone S, Dimauro C, Gambella F, Cappai MG. Industry 4.0 and precision livestock farming (PLF): An up to date overview across animal productions. Sensors (Basel). 2022;22(12):4319. DOI: 10.3390/s22124319
4.Wiseman L, Sanderson J, Zhang A, Jakku E. Farmers and their data: An examination of farmers’ reluctance to share their data through the lens of the laws impacting smart farming. NJAS: Wageningen Journal of Life Sciences. 2019;90-91(1):1-10. DOI: 10.1016/j.njas.2019.04.007
5.Wolfert S, Ge L. Cor Verdouw, Marc-Jeroen Bogaardt, Big Data in Smart Farming – A review. Agricultural Systems. 2017;153:69-80. ISSN 0308-521X. DOI: 10.1016/j.agsy.2017.01.023
6.Solà-Oriol D, Gasa J. Feeding strategies in pig production: Sows and their piglets. Animal Feed Science and Technology. 2017;233:34-52. ISSN 0377-8401. DOI: 10.1016/j.anifeedsci.2016.07.018
7.Pomar C, Remus A. Precision pig feeding: A breakthrough toward sustainability. Animal Frontiers. 2019;9(2):52-59. DOI: 10.1093/af/vfz006
8.Dourmad JY, Levasseur P, Daumer M, Hassouna M, Landrain B, Lemaire N, et al. Évaluation des rejets d’azote, phosphore, potassium, cuivre et zinc des porcs. Paris: RMT Elevages et Environnement; 2015. 26 p
9.Shaw MI, Beaulieu AD, Patience JF. Effect of diet composition on water consumption in growing pigs. Journal of Animal Science. 2006;84:3123-3132. DOI: 10.2527/jas.2005- 690
10.MARM—Ministerio de Medio Ambiente y Medio Rural y Marino. Guía de Mejores Técnicas Disponibles del sector porcino. 2010
11.Putman B, Hickman J, Bandekar P, Matlock M, Thoma G. A Retrospective Assessment of US Pork Production: 1960 to 2015. Final Report. Fayetteville, Arkansas, USA: University of Arkansas; 2018
12.You S, Liu T, Zhang M, Zhao X, Dong Y, Wu B, et al. African swine fever outbreaks in China led to gross domestic product and economic losses. Nature Food. 2021;2:802-808. DOI: 10.1038/s43016-021-00362-1
13.Gómez Y, Stygar AH, Boumans IJMM, Bokkers EAM, Pedersen LJ, Niemi JK, et al. A systematic review on validated precision livestock farming Technologies for pig Production and its Potential to assess animal welfare. Frontiers in Veterinary Science. 2021;8:660565. DOI: 10.3389/fvets.2021.660565
14.Jorquera-Chavez M, Fuentes S, Dunshea FR, Warner RD, Poblete T, Unnithan RR, et al. Using imagery and computer vision as remote monitoring methods for early detection of respiratory disease in pigs. Computers and Electronics in Agriculture. 2021;187:106283. ISSN 0168-1699. DOI: 10.1016/j.compag.2021.106283
15.Sadeghi E, Kappers C, Chiumento A, Derks M, Havinga P. Improving piglets health and well-being: A review of piglets health indicators and related sensing technologies. Smart Agricultural Technology. 2023;5:100246. ISSN 2772-3755. DOI: 10.1016/j.atech.2023.100246
16.Tuyttens FAM, Molento CFM, Benaissa S. Twelve threats of precision livestock farming (PLF) for animal welfare. Frontiers in Veterinary Science. 2022;9:889623. DOI: 10.3389/fvets.2022.889623
Written By
Robert T. Burns and Robert Spajić
Submitted: 10 January 2024Reviewed: 11 March 2024Published: 24 April 2024
Open access peer-reviewed chapter - ONLINE FIRST
