Evolution of Radio Frequency Identification (RFID) in Agricultural Cold Chain Monitoring: A Literature Review

Radio Frequency Identification (RFID) is a technology providing considerable opportunities to improve quality control for perishable foods. Over the past decade, a significant improvement in RFID application has been observed in cold chain monitoring. The aim of this paper is to, first, demonstrate the role of RFID in improving the monitoring of the agricultural products cold chain. Particular focus is placed on the specifications of RFID and its advantages, which makes its application appealing in food temperature monitoring. Second, this paper aims to provide an overview of RFID developments in cold chain monitoring. For this purpose, we conduct a review of the literature throughout 2004-2018 citing the challenges of this technology’s practical implementation in temperature monitoring of perishables, and provide the solutions presented in the literature for each limitation. This survey would be beneficial for those involved in food distribution, as it offers approaches for overcoming the limitations of RFID, making its application more advantageous.


Introduction
Supply chain management (SCM) is the "management and control of all materials and information in the logistics process from acquisition of raw materials to delivery to the end user" (Aitken, 1998). Agricultural SCM is a complicated process. Agricultural commodities have to undergo a series of operations including harvesting, threshing, winnowing, bagging, transportation, storage, processing, and exchange before they reach the market (Somashekhar et al., 2014). Advancements in information system technology have had a remarkable impact on agricultural SCM, helping it to evolve and leading to the enhancement of its logistical processes.
In an environment as complex as a supply chain, interfaces between information systems across the supply chain are necessary. With technological advancements in this area, supply chain partners can now work in close coordination to improve chain-wide efficiency and effectiveness. For the information systems to communicate proper data across the supply chain, automatic identification technologies are required. Some of the most adopted technologies include RFID, barcoding, Global Positioning System (GPS), and wireless sensor network (WSN), which aim to increase the efficiency of the supply chain network and ensure reliable information sharing. These technologies provide greater visibility by providing real-time information throughout the supply chain, allowing adaptability to market changes.
Among automated identification technologies, RFID has been central in revolutionizing agricultural supply chain networks. Within the past few years, RFID has shifted from inconspicuousness to mainstream application. RFID provides wireless identification of objects by remotely detecting electromagnetic signals. A classic RFID model has three components: a tag; a reader, which collects data; and database and information management software (Lee et al., 2006). Radio signals are emitted from the reader to activate the tag and allow the data to be received. Through the reader, communication is established between the tag and the transceiver. The RFID tag, when found in an electromagnetic zone, detects the activation signal of the reader. The reader then decrypts the data encrypted in the integrated circuit of the tag and transfers the data for processing. RFID tags can be active, passive, or semipassive. Passive and semi-passive RFID tags send their data by reflection or modulation of the electromagnetic field that was emitted by the reader ).

RFID Application in the Agricultural Supply Chain
Automated identification technology has been used for decades at different levels and in a diversity of applications in the agricultural supply chain. However, the utilization of RFID technology in various sectors of the agri-food industry has gained considerable attention in recent years.
This paper aims to demonstrate the role of RFID in improving agricultural products cold chain monitoring, and to provide an overview of developments in RFID application in monitoring the cold chain for perishables. For this purpose, we conduct a review of the literature on this application of RFID technology throughout 2014-2018, focusing on the advantages and disadvantages of its application. Finally, our study provides the solutions suggested in the literature for the various limitations of RFID application. Section two of this paper explains the research methodology. Section three gives an extensive review of previous studies on the application of RFID in monitoring the cold chain for different agricultural products throughout 2004-2018, while highlighting the advantages and disadvantages of the technology's adoption. The paper concludes by summarizing (Tables 1 and 2) the various opportunities and challenges arising from RFID application in monitoring the cold chain for agricultural products, and reviewing and presenting coping mechanisms for the challenges. Then, based on these mechanisms, we discuss the evolution of RFID application in monitoring the agricultural products cold chain.

Methodology
The main objective of this paper is to provide an overview of RFID developments in cold chain monitoring, while identifying the existing challenges of RFID implementation in monitoring the cold chain for perishable products, and distinguishing solutions to these challenges. For this purpose, we review challenges of RFID implementation in monitoring the cold chain for agricultural products cited by scholars and also provide solutions mentioned in literature to each challenge. This kind of review demonstrates the evolution of RFID in the food sector over time.
This research is a bibliographic study and the process consisted of reviewing a total of 100 scholarly studies from refereed publications (including journal articles, book chapters, and theses) from 2004, since the application of RFID in refrigerated vehicles was first proposed that year. The publications were identified through computeraided searches of renowned databases, and the articles found using keywords such as "temperature management and RFID", "cold chain monitoring and RFID", "transport and RFID", or "shelf life and RFID," and selected according to the following two criteria: (1) It must contain explanations of RFID implementation in monitoring the cold chain for various agricultural products.
(2) It must describe the advantages and disadvantages of RFID implementation in agricultural cold chain monitoring and preferably some suggestions to cope with the disadvantages.
The review is structured in two parts: (1) the advantages of RFID, which encourage its adoption in the temperature monitoring of perishables, and (2) its limitations and the suggestions for coping with these cited in the literature (where one exists).

Monitoring the Cold Chain for Perishable Foods by Means of RFID: Review of Literature
Perishable food has a considerable wastage rate. Globally, 1.3 billion tons of food are wasted every year (Buchner et al., 2012). The root cause of such a loss has been linked to insufficient temperature control (Pang et al., 2011). Low temperatures cause freezing damage, while too high a temperature has the potential to increase the rate of enzymatic and/or product spoilage (Badia-Melis et al., 2018). Fresh foods need to be stored and transported in a temperature-controlled environment to maintain or even improve their quality (Thakur & Foras, 2015). Controlling the temperature of perishable foods is beneficial for supply chain partners-it can strengthen the confidence of consumers and enhance the partners' reputation for supplying high-quality food (Kim et al., 2016;Jedermann et al., 2009). Some of the temperature monitoring systems used in the cold chain, such as strip chart recorders and temperature data loggers, which can follow the flow of products during storage and transport, are usually expensive and not automated (thus manual inspections are required), and they cannot provide a history of temperature data to which problems occurring in the cold chain can be traced. For these reasons, the use of these systems is limited to only some functions of the cold chain or to only a few types of products (Abad et al., 2009).
RFID is an emergent technology implemented across the food supply chain (Badia-Melis et al., 2016). "It can keep a record of the product's temperature history and provide accurate localized measurement results through an unbroken calibration to meet accepted standards" (Opara, 2003). RFID tags can be installed on the delivery item during the distribution process and used for automatically identifying the items based on the unique identifier assigned to each tag. By combining the information transmitted from RFID and temperature sensors, one can relate items with the environmental conditions (Michael & McCathie, 2005;Amador et al., 2009;Amador & Edmond. 2010;Kim et al., 2016).

Advantages of RFID Application in Monitoring the Cold Chain of Perishables
The use of wireless sensors including RFID in refrigerated vehicles was proposed originally by Qingshan et al. (2004). Such vehicles can host a variety of sensors to detect, identify, log, and communicate what happens during the product's journey, with the sensors monitoring the status of perishable products during transportation. One of the most prominent characteristics of RFID, which favors its application in cold chain monitoring, is the ability of RFID tags to work even in harsh conditions (such as in a temperature of minus 40 degrees centigrade) (Michael & McCathie, 2005;Abad et al., 2009).
Following the proposal of 2004, several researchers investigated the application of RFID in monitoring the cold chain for perishables and found significant advantages for RFID adoption in temperature monitoring. According to Gras (2006), implementing RFID in the transport system offers the potential to manage issues in the supply chain through automatic identification. In other words, RFID does not entail manual scanning, and provides data remotely and continuously (see Table 1 for references). This characteristic leads to a decrease in labor cost and alleviates the human error that occurs with manual systems (Table 1).
On the other hand, automatic identification and remote data collection by RFID provide the possibility of both saving time and real-time quality control. Hundreds of RFID tags can be downloaded into and read from a computer at one time (Table 1), which could lead to a decrease in shipment confirmation time (Gras, 2006). Furthermore, while the large memory of RFID systems permits the collection of huge amounts of temperature data, these data can be recovered and read quickly (Table 1). This permits the collection of real-time information on the quality status of the product, and favors the application of RFID in food cold chain monitoring, which requires real-time decision-making (Table 1). Several researchers have taken advantage of the real-time temperature data provided by RFID.  proposed an innovative planning model to optimize perishable product allocations from manufacturers to distributors and retailers. The 2008 Olympic Games in Beijing also benefited from RFID application in the food sector. The temperature of the food storage compartments of all vehicles transporting food served at the Games was real-time monitored, which ensured the safety of the transported foods (Wu et al., 2010). Chen and Shaw (2011) and Zhang et al. (2018) benefited from the advantage of RFID in providing real-time temperature data, and proposed a real-time temperature control mechanism for agricultural products. Trebar et al. (2015) tapped into the potential of RFID to define the optimal cooling materials for fish packaging. Besides this, Zhou and Chakrabartty (2017) benefited from the advantage offered by RFID in providing continuous information for their study on using the continuous integration and storage of information as a time-temperature indicator.
Another advantage of the RFID system is the technology's quick instrumentation (Amador et al., 2009).
In addition, continuous temperature control and close temperature readings of perishables (by embedding temperature sensors within the loads) (Ruiz-Garcia et al., 2008) enable RFID to trace any problem arising throughout the product's journey (Ogasawara & Yamasaki, 2006;Erdem et al., 2008). In addition, RFID provides a means to identify a defective product, thus helping to quicken reactions to any quality problem (Regattieri et al., 2007;Trebar et al., 2011) and to prevent the recall of products and associated costs (Kumar & Budin, 2006). The ability of RFID tags to provide non-cumulative information can be considered another of RFID's advantages (Lorite et al., 2017), and this, accompanied by RFID's ability to provide real-time information, can lead to efficient risk management in cold chains, as both are beneficial for risk identification and response steps (Kim et al., 2016).
The real-time and continuous information provided by the RFID system can be also used in determining the freshness of perishables. According to Grunow and Piramuthu (2013), RFID provides an advantage in estimating the shelf life of perishables, as it is able to provide more information on the quality of products than the expiration date provides. Hertog (2006) Some authors argue that RFID could contribute to an increase in transparency in the agricultural supply chain (Table 1). Focusing on the role of temperature indicators in supply chain transparency, Lorite et al. (2017) argued that time-temperature indicators provide only a cumulative indication of cargo temperatures. With the aim of providing non-cumulative information, the authors suggested the integration of RFID into a simple critical temperature indicator (CTI).
RFID application also provides some economic benefits. Alleviating the cost of a PC connection (Note 1). (Ogassawara & Yamasaki, 2006;Tingman et al., 2010), increasing sales by reducing product loss (Estrada-Flores & Tanner, 2008;, reducing labour cost, reducing inventory, and improving return on investment (Estrada-Flores & Tanner, 2008) are some of the direct benefits of RFID implementation in perishable product supply chains. The reusability of tags and readers, which makes the application of RFID in the food industry costeffective, adds to the direct benefits of RFID application in temperature monitoring (Ting, 2013;. On the other hand, improved consumer satisfaction can be considered an example of the indirect and nonfinancial benefits of the RFID system, and can increase direct benefits later . Some authors have investigated the economic benefits of RFID application in cold chain monitoring. Hong et al. (2011) confirmed a positive profit from RFID application in a chain of convenience stores in Taiwan after years 3 and 4. The results of a study by Ketzenburg et al. (2015) indicated that significant value of information could be generated by the application of RFID in monitoring the cold chain for perishables through decreased product spoilage.

Limitations of RFID Application in Monitoring the Cold Chain of Perishables and Their Solutions
The vast potential of RFID in cold chain monitoring has led researchers to take advantage of the technology's benefits in their studies and to strive to find solutions for its limitations. Over time, these efforts have led to the technology's evolution. In this study, we categorize the limitations found in the literature as organizational, technical, and those related to security and privacy. We conduct a literature review for each category of RFID limitations, and provide the solutions outlined in the literature for each limitation.

Organizational Limitations
Scholars have cited the organizational challenges that limit RFID adoption in agricultural cold chain monitoring, which include a lack of skilled personnel (Asif, 2005;Attaran, 2009;Raab et al., 2011;Ruiz-Garcia & Lunadei, 2011), resistance to change, a lack of top management support, a lack of system integration (Attaran, 2009), and poor consumer adoption (Estrada-Flores & Tanner, 2008). Estrada-Flores and Tanner (2008) suggested that RFIDenabled refrigerators with functions such as sending a shopping list to the owner's cell phone could make the benefits of RFID more evident for consumers and thus improve consumer adoption of the RFID system.

Technical Limitations
Besides organizational challenges, the technological limitations of RFID adoption have been mentioned in the literature. Some of these are related to the structure of the RFID system. Each component of the RFID system is subject to several limitations; we categorize these based on the challenges related to each component.
Previous RFID applications in agricultural cold chain monitoring have revealed some limitations related to RFID tags. First of all, multiple types of RFID sensors (e.g., semi-passive temperature sensors and passive tags) need different architectures, as they require different environments for identification (Delen et al., 2011). Sensing diversity has also been cited as a limitation of RFID application; i.e., usually, RFID technology will be limited by a single sensor, such as temperature Badia-Melis et al., 2018). In addition, RFID tags are not able to initiate communication . Tagging level can also affect the efficiency of cold chain monitoring. RFID tagging can be implemented at the pallet level, case level, and item level. Pallet-level tagging is a remarkable challenge when different products are transported together, since different products show different sensitivity to temperature (Raab et al., 2011).  looked to identify the optimal temperature setting for an environment including chilled and chill-sensitive with different temperatures. For this purpose, they proposed a smart cold chain management (SCCM) framework, in which an RFID tag is attached to the stock keeping unit (SKU) of each product and a WSN system is deployed to each cold chain facility. On the other hand, Ruiz-Garcia and Lunadei (2010) pointed out the advantage of item-level tagging in providing more detailed information, as even within the same pallet or case, different items can experience different temperatures (depending on their distance from the cooling unit or from the walls, etc.) and non-uniform ripening. Thus, itemlevel tagging could be beneficial in preventing errors in decision-making that has been based on temperatures affecting only part of the load (Grunow & Piramuthu, 2013).
Tag cost is a major restrictive factor in the implementation of RFID technology in the agri-food industry ( Table  2). Ruiz-Garcia and Lunadei (2011) pointed out that the average cost of RFID tagging is considered overly expensive for low-value agricultural products such as fruits and vegetables (Regattieri et al., 2007;Chunxia et al., 2009;Ruiz-Garcia & Lunadei, 2011). Several researchers have provided solutions for coping with this limitation. Bhattacharyya et al. (2010) discussed the potential for RFID tags to be used as low-cost sensors in monitoring the cold chain for agricultural products. According to the authors, this could be done through plotting a change in a physical parameter of interest to a controlled change in the electrical properties of the RFID tag antenna. Some authors, such as Jedermann and Lang (2007), Amador and Edmond (2010), and Trebar et al. (2013), have suggested reducing the number of tags. In their study, Jedermann and Lang (2007) reduced the number of tags in the Intelligent Container, while Amador and Edmond (2010) suggested reducing the number of tags by installing them in a location where the temperature is the middle point of an interval. Trebar et al. (2013) suggested reusable smart tags, called RFID-TLs. In another study, Chen et al. (2014) proposed the application of a new type of RFID named 2G-RFID-Sys (second-generation RFID systems). In the 2G-RFID-Sys, a dynamic rule encoding "mobile code" is stored in the tag. One of the main advantages of this system is that it replaces semi-passive tags with passive tags, hence reducing the cost. Several challenges concerning the RFID reader were also found in the literature; among these was the readability of RFID tags. Some authors mentioned the low reading range of RFID tags as a disadvantage of the technology (Table 2). According to Kumar et al. (2009), most low-frequency RFID systems have an operating range of about 1 meter, and ultra-high frequency systems extend the operating range to 3 to 4 meters. However, Badia-Melis et al. (2015) demonstrated that a combination of a WSN and RFID could solve the problem of the tags' low reading range, as a WSN offers a long reading range. The RFID tags' reading range can be affected by different factors. Gras (2006) stated, "Material mounted transponders could lead to different reading ranges." Another critical issue is the sensitivity of RFID tags' readability to environmental factors such as dust, moisture, extreme temperatures, the presence of metallic objects, and the high-water content of tagged materials (Table 2). Using RFID tags for agri-food products that have a high-water content and are usually transported in metal containers, such as fruits and vegetables, could be challenging (Angeles, 2005; Ruiz-Garcia, 2008). This limitation results in the generation of inaccurate spoilage predictions and increased food spoilage (Bedia melis et al., 2018). To cope with this challenge,  suggested the use of non-metallic containers for transporting perishables by air. Li et al. (2012) suggested the link quality indicator be measured in a cargo container filled with watered Styrofoam boxes (Note 2) to ensure excellent reading efficiency of the WSN. Several authors tested the RFID antenna position to achieve full RFID tag readability and strove to find a frequency/configuration that would allow real-time reading of the temperature in a shipment of perishable products using RFID. Laniel et al. (2009) concluded a significantly higher performance at the 433 MHz level (compared to 915 MHz and 2.45 GHz). Moreover, an RFID antenna positioned at the front of a container of perishable products delivered slightly better results than the one in the middle of the ceiling. Laniel and Émond (2010) concluded that canned vegetables provide the best RFID tag reading, followed by fresh meat and frozen bread. The results of their study demonstrated that radio waves at 915MGH are absorbed by high water and ice contents. However, spacing between metal surfaces and tags can allow excellent RFID readability even if the radio frequency reflects on metallic surfaces. The authors suggested that RFID tags placed on the top of a pallet have a higher chance of producing a successful read. The results of a study by  showed that antenna position affects RFID readability at 915 MHz, while 433 MHz demonstrated 100% readability in all antenna positions. Trebar et al. (2013) concluded that a high frequency is considered the best solution for the integration of foods with water conditions. Accuracy is also a concern with the RFID reader. Significant factors affecting the reader's accuracy are the visibility of tags to the reader (Asif, 2005;Kumar et al., 2009) and harsh conditions in the cold chain. Although in harsh conditions RFID tags might continue to work, readers fail to register tagged pallets (Estrada-Floes & Tanner, 2008;. It is worth mentioning that the probe-less RFID system fails to provide an accurate temperature reading in critical corners of a pallet (Amador et al., 2009). Amador et al. (2009) proposed the use of RFID semi-passive tags with a probe for tracking low temperature abuse in the transportation and refrigerated storage of products.
Data management can be considered a limitation of RFID information management software. As mentioned earlier, RFID enables continuous temperature data collection. Although this characteristic is very beneficial in cold chain monitoring, managing the huge amount of data produced by RFID data loggers is a restrictive factor in its adoption (Table 2). Moreover, a large volume of data leads to high hardware cost (Jedermann et al., 2009). However, Badia-Melis et al. (2015) and Badia-Melis et al. (2016) stated that in the case of creating a uniform mesh of temperature nodes with many data points, which is necessary for 3D representation, a large volume of data is not a limitation. Fortunately, several suggestions to combat this limitation were found in the literature. Referring to the patent of Martin and Grabiner (2007), who proposed the integration of RFID tags with a visually readable environmental condition exposure indicator, Jedermann and Lang (2007) and Jedermann et al. (2008) suggested replacing huge amounts of temperature data with a temperature threshold checker and a quality indicator, respectively. In another study, Jedermann et al. (2009) stated that the solution to this limitation lies in implementing a decentralized data management system in which temperature data are pre-processed close to their point of origin by intelligent systems. Finally, Raab et al. (2011) proposed that temperature monitoring system providers consider reducing large amounts of data into a summarized form to enable rapid management decision-making.
Real-time data transfer to the RFID interface is considered another challenge of RFID database and information management software. Several authors argued that previous methods of temperature monitoring using RFID tags with integrated temperature sensors (Jedermman & Lang, 2007;Thakur et al., 2011) could not retrieve information in real time when the product was in transit (Note 3) (Thakur & Foras, 2015). Integrating RFID with other technologies, these authors improved ancient methods of temperature monitoring to achieve a real-time monitoring system (Zhang et al., 2009;Wang et al., 2010;Thakur & Foras, 2015).
Other technical challenges besides those related to the structure of the RFID system were found in the literature. A lack of both uniform standards and a non-uniform calibration procedure between manufactures and models remain challenges of RFID implementation in cold chain monitoring (Table 2). Although EPCglobal and the International Standards Organization (ISO) worked to develop international standards for RFID technologies in the UHF (ultra-high frequency) spectrum, the standards put forth by these organizations were not completely compatible with each other (Chunxia et al., 2009). McMeekin et al. (2006 proposed using the EAN.UCC system to solve the problem of non-uniform standards. Jedermann et al. (2006) pointed out the lack of a suitable quality indicator as a technical challenge of RFID application in cold chain monitoring. Since commonly used quality indicators-including firmness, starch and sugar content, taste, and color-are not suitable for automated supervision because the products have to be unpacked and handled manually, the authors looked to improve these indicators by combining a WSN and RFID.
By incorporating a factor directly associated with the internal metabolic process, their quality indicator measures the quality of the fruit itself.
RFID also has high energy consumption, which can be solved using an ultra-low power complementary metal oxide semiconductor (CMOS) with room-temperature targeting (Law et al., 2010).
The high cost of RFID implementation and small value added of agricultural products have led researchers to challenge the economic benefits of RFID technology in the food industry (Table 2). An RFID system must be customized to its specific working environment and application process (Chunxia et al., 2009), which is costly. Given the high cost of RFID application, Chunxia et al. (2009) considered return on investment an important issue in RFID application in the vegetable supply chain with low value added. Delen et al. (2011) justified the return on investment in their application by an increase in quality, which would promote sales, and the reduced risk of losing a customer due to poor-quality products.

Limitations Related to Security and Privacy
Security and privacy concerns represent a significant limitation of RFID implementation (Table 2). Consumers are concerned about automatic tracking of their movements and buying habits. An unauthorized user of an RFID system can also scan tags to access private information such as shipment and inventory details (Kumar et al., 2009). Langheinrich (2009) stated that because of the tags' low computational powers, a simple system must be configured to prevent tampering with data and to protect privacy.  Raab et al., 2011;Grunow and Piramuthu, 2013;  Pallet-level tagging  Ruiz-Garcia and Lunadei (2010) proposed item-level tagging  Grunow and Piramuthu (2013) proposed the application of itemlevel RFID tags  Aung and Chang (2014) decided the optimal target temperature in a multi-commodity cold storage management   Disability of a tag to initiate communication Badia-Melis et al., 2018  Lack of sensing diversity Asif, 2005;Michael and McCathie, 2005;Narsing, 2005;Jedermann and Lang, 2007;Regattieri et al., 2007;Erdem et al., 2008;Estrada-Flores and Tanner, 2008;Abad et al., 2009;Chunxia et al., 2009;Amador and Edmond, 2010;Bhattacharyya et al., 2010;Zou et al., 2014;Sima et al., 2017;Badia-Melis et al., 2018  High cost of tags  Jedermann and Lang (2007) Angeles, 2005, Asif, 2005Dobkin and Weigand, 2005;Narsing, 2005;Regattieri et al., 2007;Ruiz-Garcia, 2008;Laniel et al., 2009;Laniel and Émond, 2010;Amador and Edmond, 2010;Ruiz-Garcia and Lunadei, 2010;Raab et  Limitations related to database and information management software Jedermann and Lang, 2007;Martin and Grabiner, 2007;Erdem et al., 2008;Estrada-Flores and Tanner, 2008;Jedermann et al., 2008;Jedermann et al., 2009;Ruiz-Garcia and Lunadei, 2010;Raab et al., 2011;Ruiz-Garcia and Lunadei, 2011;Kumari et al., 2015  Data management  Jedermann and Lang (2007) Chunxia et al., 2009;Yan and Lee, 2009;Delen et al., 2011;Raab et al., 2011;Ruiz-Garcia and Lunadei, 2011;Kumari et al., 2015;Thakur and Foras., 2015  High cost of implementation  Delen et al. (2011) justified the return on investment in their application by a reduction in spoiled products, an increase in quality that would promote sales, and the reduced risk of losing a customer due to poor quality products Asif, 2005;Michael and McCathie, 2005;Kumar and Budin, 2006;Kumar et al., 2009;Langheinrich, 2009;Yan and Lee, 2009;Ruiz-Garcia and Lunadei, 2010  Security and privacy  Langheinrich (2009) stated that with the low computational powers available within the tags, a system to prevent tampering with data to protect privacy must be simple

Discussion and Conclusion
From 2004 to 2018, the application of RFID technology in agriculture increased rapidly. Although RFID brings new features to traditional temperature monitoring when adopted in cold chain monitoring, some challenges limit the technology's application.
The main reasons for the enthusiasm for RFID adoption in cold chain monitoring are its ability to integrate temperature sensors, provide real-time information, and work in the harsh conditions of cold chains. Table 1 presents the advantages of RFID application in monitoring the cold chain for agricultural products cited by scholars during 2004-2018. These advantages are categorized into: time saving, data collection, risk management, structure (of the RFID system) and economic benefits. The significant advantages of RFID application in temperature monitoring have led researchers to look for solutions to the limitations of the application of this technology in the cold chain of perishables. We categorized the challenges of RFID implementation in cold chain monitoring according to Nambiar (2009) into organizational, technical, and those related to security and privacy; technical challenges were further categorized into limitations related to RFID tags, readers, database and information management software, and others ( Table 2). The solutions offered in the literature for these limitations are presented in Table 2. Given the long period (14 years) that RFID has now been adopted in cold chain monitoring, organizational limitations such as a lack of skilled personnel are no longer an issue, nor is a lack of management support, as the risks associated with new technology adoption occurs in the first stages of its emergence, and the significant benefits offered by RFID earn management support for its adoption and provides sufficient incentive to the manager to train personnel. Technical challenges were the most cited concerns with RFID implementation in cold chain monitoring during 2004-2018 (Table 2); while solutions were found for some of these, some technical challenges (environmental requirements of multiple types of sensors, sensing diversity, tags' failure to initiate communication, and reader accuracy) remain. Reader accuracy is a serious challenge, since it can be affected by the harsh conditions of the cold chain, leading to error in decision-making. Finally, the unauthorized traceability of buyers and venders remains an issue.
This literature review outlines the evolution of RFID application in agricultural products cold chain monitoring. Several solutions have been found over time for fundamental issues arising from RFID application in temperature monitoring. Thus, given the technology's significant advantages, it could be considered a viable method of temperature control providing both direct and indirect benefits to the involved parties. The application of RFID technology is becoming more and more beneficial, as tag cost has been steadily declining recently and has reached the threshold value, making RFID application profitable (RFID journal, 2018). Moreover, as RFID infrastructure is now built across the entire food supply chain (cultivating, processing, storage, transportation, and retailing), the resulting economies of scale leads to more benefits overall for the agricultural industry.