1. Introduction
Food safety is an essential and foundational right for customers. Despite several significant advancements in food science and safety, foodborne infections remain a serious public health concern around the world. The Centers for Disease Control and Prevention (CDC) reported that around 48 million people die due to contaminated food ingestion every year in the United States. As a result, 128,000 of them are hospitalized, and 3000 die [
1].
According to research, the financial impact of foodborne diseases in the United States is estimated to be around
$55.5 billion [
2]. This financial cost is incurred due to hospitalizations, decreased productivity, economic losses, and a variety of other factors. Additionally, these official estimates exclude other costs, such as the life-long health repercussions of foodborne infections. In January 2013, the discovery of horse DNA in frozen beef burgers attracted attention to the issue of meat adulteration [
3]. Horsemeat is utilized in place of beef due to its lower cost of production.
Table 1 presents various foodborne illness outbreaks that occurred around the world, as analyzed in the work of Gourama [
4].
Food safety and food security are connected concepts that have a significant impact on the quality of human life, and both of these areas are influenced by a variety of external circumstances. These concepts are of great importance for the health of consumers, and experts in various fields are constantly trying to cope with any challenges met across the food supply chain. New and emerging technologies such as Internet of Things (IoT), blockchain and tiny machine learning (TinyML) seem to be the tools capable of enhancing any expert attempts of ensuring the two complementing elements in consideration.
More precisely, food safety is a process that involves a variety of activities completed by people who come into contact with various forms of food at various development and operation stages throughout the world, in order to achieve a fixed food safety standard that meets both general and specific requirements. The potential problems associated with the world’s population growth underlined the vital role of all actors in resolving food safety issues, including producers, distributors, consumers, government agencies, scientists, and medical experts. Food safety culture is mirrored in an organization’s technological and managerial aspects, as well as in its personnel and working environments. Food safety should include a variety of management strategies, including the regular monitoring and surveillance of food production, in order to improve public health and avoid foodborne infections [
12,
13,
14].
In contrast, the notion of food security ensures that everyone has access to appropriate, secure, and healthy food in order to maintain a healthy and active lifestyle. The determinants of IoT that facilitate data generation and collection, such as electronic control, smart contracts, policy enhancement, and the use of radiofrequency identification (RFID), are argued to be critical enablers for a motivated food security system, food safety, and environmental sustainability [
15]. Food waste reduction and appropriate waste management mitigate the negative environmental impacts of food waste, preserve economic resources, and promote food security [
16]. Blockchain technology is a promising approach for reducing food loss, increasing transparency, stakeholder confidence, and food security [
16].
One of the most pressing problems nowadays is the provenance of the food and food-related items that people consume, owing to several food scares and the globalization of food markets, which has resulted in food movement between nations and continents. Food product documents and paper trails may have mistakes or even be faked by criminals, resulting in inaccurate or insufficient information on product labeling For instance, Halal certification, Islamic values, and Halal food safety are among the most critical markers of a Halal sustainable food supply chain [
12]. However, an innovative system for ensuring the “Halalness” of items, such as the Halal label, has recently lost the faith of Muslim customers due to a rash of food fraud instances [
17]. Thus, establishing traceability and visibility via the use of proper methodologies is critical to fostering confidence among stakeholders [
18].
Globalization has altered the food system, and consumer demographics and behavior are changing. A sizable segment of the population is aging or becoming immunocompromised. Consumers are requesting a greater variety of fresh fruits and vegetables, as well as minimally processed goods [
19]. In Europe, people are ready to pay a premium for high-quality items that include information on the product’s origin, species, and variety. Food scientists present a considerable difficulty in validating labeling compliance in a way that is acceptable to the whole food business and the customer [
20]. The only way to overcome these obstacles is for stakeholders across the food system to be willing to experiment in new and innovative ways.
Furthermore, temperature [
19] has always been an element that attracts great interest from researchers and food technologists because it demonstrates a wide variety of interactions between microbes and food matrices. The temperature significantly affects the growth and inactivation rates of pathogenic bacterial infections. Foodborne infections are generally mesophilic, flourishing between 20 and 45 °C.
For
B. cereus spores, a slight temperature increase from 2 to 8 °C can result in a tremendous expansion of up to 10
3 B. cereus/mL in around nine days [
21]. This slight temperature increase may occur during the transportation of refrigerated goods and may also correspond with the temperature of commercial refrigerators. This is one of the first indications of why IoT devices must be utilized for temperature monitoring, and write this information to a system such as the blockchain. If a product is affected or partially affected by the spores in discussion, having this information is crucial for the recall process, since the product is not safe for consumption according to the European Commission regulation No 2073/2005 [
22]
The cooling rates of various foods can be used to assess whether microbes require cold shock proteins or are able to adapt to changes in the microenvironment. Inadequate chilling periods can promote microbe development, particularly those whose spores survived the cooking process [
23]. Spore-formers such as Clostridium botulinum, Clostridium perfringens, and Bacillus cereus can germinate and proliferate rapidly when foods are not properly chilled, resulting in foodborne disease [
24].
Microorganisms’ capacity to develop and survive within a food product is governed by the food’s composition and environment, the processing parameters used, and the storage conditions used during the food’s shelf life. The term “intrinsic factors” refers to the properties of the food matrix. In comparison, extrinsic variables are the characteristics of the surrounding environment, particularly during processing and storage. From the raw components to the finished product during storage, the food matrix and ambient circumstances undergo several changes, all of which may contribute to product development [
25].
Some works [
26,
27,
28] analyze the need for records to provide evidence to buyers and regulators that the product followed the correct procedures before reaching the customer. Other studies such as [
29,
30,
31] discuss the significance of food traceability and how it may assist in solving food safety problems and enhance supply chain performance, particularly in terms of sustainability and customer transparency. Additionally, it is noted that, in order to retain public confidence, comprehensive and sustainable tracking and recall procedures must be built and maintained. This is now possible thanks to advancements in technology such as blockchain and IoT sensors.
Blockchain technology can be identified as a way to improve existing systems by boosting their transparency and traceability, hence regaining consumer trust. Blockchain technology may be used to construct food supply chain systems that are more transparent, traceable, and sustainable [
32]. Distributed ledger technologies improve transparency and traceability in the agricultural sector’s information flow. The technology possesses the following three major capabilities: The primary offerings are as follows: (a) ensuring the authenticity of the product by tracing its provenance and recording all transactions; (b) enabling secure, effortless, and real-time payments; and (c) facilitating better production and marketing decisions through accurate data monitoring and storage. Additionally, food products would be traceable along the supply chain using the blockchain’s unique IDs. Food waste may be prevented with information about growth conditions and expiration dates, and the immutable record of items and transactions can help avoid fraud and foodborne disease [
33].
Moreover, blockchain technology may be used to enhance the following functions in food supply chains [
34]:
Track the flow of goods along with the supply chain
Logistics tracking, e.g., orders, receipts, and shipping alerts
Attributing certifications and characteristics to products
Connecting items to their serial numbers or digital tags
Sharing vital information across the product’s assembly, distribution, and maintenance
As stated above, another important factor is the utilization of IoT devices for monitoring environmental or other critical factor metrics to ensure the quality of food products. However, these devices are not built with security mechanisms and are easy to exploit through the different communication protocols that they use to connect to third-party entities, such as other devices or the cloud. In order to be able to create an end-to-end system capable of providing unaltered information, scientists must come up with ways of securing them or developing new robust and resilient systems. Nowadays, typical IoT devices utilize several different kinds of sensors that are easy to be tempered from malicious actors. TinyML is an emerging technology that can operate in constrained hardware and provide intelligent results by running machine learning (ML) locally on edge. Additionally, since there is now the ability to run complex models locally, Anomaly detection systems can be developed to detect and report when abnormalities are discovered across the supply chain. These abnormalities could be a result of a spoiled food product, faulty machine operation, or attempts to tamper with monitoring devices.
Developing secure and robust systems for supporting critical operations, such as supply chain monitoring, is of crucial importance. The use of blockchain technology is one step forward in the right direction, as it ensures the security of the system upon which end nodes interact. The secure operation of the end nodes of the systems through the integration of proper security mechanisms is the second step we are required to take. It will ensure (to the extent that this is feasible) the normal operation of the end nodes and the validity of the data input for the system as a whole. Systems are as secure at their least secure node, and in order to develop fully secure systems, we need to improve security for all their parts. It is essential for the supply chain domain to provide both a secure blockchain-based backbone and a lightweight security mechanism for edge devices.
The contribution of the present paper is two-folded:
It introduces a blockchain-based system that can be employed to store data across the food supply chain, from farm to grocery. The system ensures that data tampering is infeasible and delivers end-to-end transparency for all actors in the food supply chain.
It proposes a security mechanism, based upon the emerging technology of TinyML, that can be integrated as a security control to the edge devices used for monitoring purposes. This mechanism is based on a lightweight anomaly detection approach for the monitored data and is capable of identifying cases where malicious actors attempt to exploit or tamper with the devices.
Combining these two complementary mechanisms increases the citizens’ trust concerning data presented to them regarding consumed food products.
The rest of this paper is organized as follows.
Section 2 presents previous work regarding blockchain-based systems utilized in the food sector.
Section 3 provides a brief overview of the main requirements of the food industry.
Section 4 describes the three distinct subdomains of the food industry, agriculture, livestock, and fishery.
Section 5 introduces the Blockchain technology.
Section 6 introduces the technology of TinyML, it’s main advantages and requirements, and TinyML-based systems regarding agriculture, machine failure prediction and device’s security.
Section 7 describes the overall concept, challenges and thoughts behind the system proposed in this work
Section 8 analyzes the blockhain-based system and
Section 9 presents an example of a system capable of identifying anomalies regarding the monitoring devices required across the food supply chain. Finally,
Section 10 concludes with an analytical overview and results from the experiments conducted and presents future plans.
Figure 1 depicts the organization of the sections and the overall thoughts behind the paper. It starts with introducing challenges and issues met across the food supply chain, presents food industry’s key components and subdomains, introduces the two emerging technologies from which the blockchain-based system and the anomaly detection device are inspired and then presents them and finally concludes with comprehensive analysis and presentation of the results.
2. Related Work
According to the literature, the usage of blockchain technology in applications linked to quality assurance and transparency in the food sector has increased significantly in recent years. All of the efforts outlined are not mature enough or may not completely document how the blockchain is used, nor do they address specific issues or restrictions. Conversely, the sheer volume of scholarly publications over a short period reveals a definite trend toward integrating blockchain technology into existing food sector procedures. This trend is primarily due to the compatibility of the characteristics offered by blockchain systems with the issues confronting the particular industry. The following paragraphs provide an overview of various blockchain-based systems for the food supply chain.
To address China’s rising food safety concern, a team presented a supply chain traceability system [
35] built on blockchain and RFID (radio-frequency identification) technologies. The system can be used to collect data and manage information for all stakeholders, using a “from farm to fork” approach. The use of RFID tags, which are typically found on product packaging, enables the presentation of numerous agri-food product features to the consumer, such as the product’s name, variety, origin, fertilization state, and pesticide usage. The system collects, circulates, and shares data using RFID technology, while blockchain technology ensures the integrity of shared information. The system is suggested to be applied to the primary food industry areas of fruits, vegetables, and meats.
The authors of work [
36] provided an end-to-end solution for a blockchain-based agri-food supply chain. They have presented comprehensive details such as traceability and delivery regarding the cases explored on the proposed system in consideration. Additionally, they have thoroughly researched and evaluated the efficiency of smart contracts to ensure that the offered solution is both accurate and efficient. The reputation system is intended to sustain the agri-food supply chain’s authenticity and product quality standards. Furthermore, since these transactions are based on blockchain technology, they maintain the immutability and authenticity of the transactions. The system in consideration requires a particular quantity of gas to deploy and execute smart contracts, as demonstrated by the simulations provided. Regarding authors’ future plans, they intend to implement methods such as refunds into the trade of agri-food items. Moreover, since the reputation system archives reviews from end-users that may be skewed or fraudulent, an additional system capable of detecting fraudulent reviews is intended to be added.
Another team introduced KRanTi [
37], a blockchain-based system for the Agriculture food supply chain (AFSC) to assist in resolving production tracking and efficiency issues, as well as making the system more resilient and transparent among users. The system makes use of Ethereum to track transactions between stakeholders and to ensure consistency by maintaining a record of the score granted to the former stakeholder. Additionally, KRanTi offers farmers a special credit-based program that enables them to build funds for superior agriculture-related items. The authors provide tests, compare, and analyze various information such as bandwidth, gas usage, and data storage cost. Additionally, some limitations are mentioned, mainly regarding the overall cost, and one of their plans is the implementation of an artificial intelligence (AI) system to forecast system abnormalities.
A research team was tasked with strengthening public trust in the food supply system. Authors built their system on cutting-edge technologies such as blockchain and IoT while taking the hazard analysis and critical control points (HACCP) methodical approach seriously. The system is based on BigchainDB, a distributed database system that resembles a blockchain, and utilizes a variety of sensor and networking technologies to efficiently gather and transfer the essential data. According to preset access control rules, all supply chain participants can interact with the system by adding, retrieving, or editing data stored in the database. RFID tags are used to label products, which are then immediately associated with a virtual identity. This permits the identification and retrieval of information about each product throughout its existence in the supply chain [
38].
In this work [
39], a unique architecture that leverages IoT and blockchain technologies to increase transparency throughout the food supply chain is presented. Each food product is assigned a unique identity via the use of an RFID tag. Each action associated with this particular shipment is logged to a blockchain system via sensors deployed at all crucial points across the supply chain, resulting in a tamper-proof digital history. Consumers and retailers can access the public ledger at any time to receive product information. All of this critical information can be used to update the shelf life and conduct targeted recalls. Although only one sensor was included in the suggested system, other sensors such as moisture and light sensors might be included depending on the packed food product. Additionally, the authors did a security analysis, which revealed that the validation of a fabricated block becomes even more challenging with a more significant number of nodes and numerous consensus stages, which can be enhanced by improving hardware security.
World Wide Fund for Nature, a non-profit organization, has embarked on a blockchain supply chain project to establish a transparent and traceable environment for the fresh and frozen tuna supply chains [
40]. The pilot was explicitly designed for tuna captured in a longline fishery in Fiji. RFID and QR codes were utilized to collect data at various points along the supply chain. Each fish that lands on a fishing vessel can be traced by attaching a tag to the fish prior to its placement in the hold. This one-of-a-kind tag could be attached to the fish and automatically registered at several types of equipment located on the vessel, at the dock, and in the processing plant. Internet access was necessary to transmit and record the tagged fish data as digital assets on the blockchain. As fish are tagged, a dedicated mobile application records vital data on a mobile device, which digitally certifies the record. Once internet connectivity is established, the software immediately uploads recorded data to the proper servers, where it is then stored in the blockchain. When fishers return to the port, each tagged piece of fish is checked and then delivered to a processing plant. The tag remains attached to the fish throughout the process to ensure the fish’s history is not lost or altered.
Provenance [
41] collaborated with fishermen to develop a “first-mile” registration system. The use of blockchain and smart monitoring to supervise the fishing activity of fishers who meet particular social sustainability requirements was pioneered. Fishers could issue their catch as a new blockchain asset via an SMS message. Each object is subsequently assigned a unique permanent ID. When tangible transactions between fishermen and retailers occur, they are also recorded in the blockchain. The item credited to the fishers is linked to the suppliers, and the prior owner’s information is also maintained for backward tracing purposes. Additionally, Provenance presented applications for customers that enable them to obtain necessary information about the goods. Smart stickers equipped with NFC are employed, and when scanned, the consumer may assess the product’s transaction history from the sea to the store. Additionally, item monitoring might be extended to dining establishments, informing patrons of accessible information about items and ingredients. Provenance’s objective is to provide a solution for data interoperability and item tracking in a highly robust yet accessible format.
The work [
42] presents a comprehensive concept of a blockchain-based agrifood supply chain traceability system in this, along with a prototype implementation. The implementation of blockchain technology reduces the requirement for supply chain participants to rely on a single organization needed to handle supply chain operations and preserve traceability data. Additionally, this fully distributed strategy eliminates the limitations of scalability as well as a single point of failure. The suggested solution enables the automation of supply chain management procedures and the secure and persistent storage of traceability information. Additionally, the ability to add rules at runtime enables the flexible development of product-specific quality control procedures. Ultimately, the system offers an overview of the many stages of harvesting, processing, and distribution through which quantities of product pass, allowing for the reconstruction of each batch’s complete lifecycle and the acquisition of origin details.
Additionally, a blockchain-based system was proposed for tracing the origin of chicken products. The technique was deployed in Tien Giang province, Vietnam, on chicken farms. Numerous companies took part, including poultry farms, veterinary business agencies, and shops. The suggested system was written in PHP, and in addition to the usage of Blockchain technology, QR codes were included to allow consumers to search for information about chicken products stored in the system. Farmers and stakeholders were optimistic about participating in the experiment, and the early findings were encouraging [
43].
The authors of the study [
44] present FoodSQRBlock, a framework based on blockchain technology that digitizes food production data and makes it easily accessible, traceable, and verifiable by consumers and suppliers through the use of QR codes to integrate the information. Additionally, they used Google Cloud Platform to simulate a real-world food production scenario, employing milk and pumpkins as representations of products from real farms in the United Kingdom. Experiments demonstrate the practicality and scalability of implementing FoodSQRBlock in the cloud.
Additional blockchain-based systems for the food supply chain are discussed in the work [
45], together with the subdomains into which they are incorporated, the degree of implementation, the blockchain system chosen, and the type of blockchain access.
Table 2 showcases various blockchain-based systems proposed by the scientific community, among with the subdomain they are utilized in, the food industry requirements they are capable to enhance, and finally, their implemented technologies.
8. Blockchain Based System
The concept is based on the idea of representing food ingredients as tokens. A new token is minted for any new batch of ingredient that is produced. The ownership of such tokens is closely monitored and recorded. Apart from token transfers, the system allows for token splits and token packaging into a final food product.
Let’s assume that we aim at monitoring a food supply chain based on agricultural products. In the proposed scenario, five main actors may be identified. First of all, is the farmer, who has the ability to create a token, where a token represents an ingredient in the real world. After the production, the ingredients have to move to the factory. Therefore, another actor responsible for the transportation is present. During the transportation, the farmer has to transfer the token, and the transport actor has to receive the token. Then, when it is delivered to the factory, again the same process occurs, where the transport actor transfers the token to the factory and the factory receives it. The last actor on this chain is the grocery store, where again, has to receive the final product, which is a collection of tokens in our system. The final user of this system, which is not related to the above chain, but can and has to be able to view it, is the consumer. The consumer, by using the mobile phone, can scan a QR code on the final product, which will create a view of all the transactions and the actors involved from the farmer to the grocery store. The entire process is depicted in
Figure 3.
All of the above actors and actions are described in the following paragraphs, where an overview of the backend and the interface of the proposed system are analyzed.
8.1. Smart Contracts
The core of the system consists of two smart contracts. The former is responsible for the governance of the users and the latter is responsible to maintain all the information on the chain. The first contract, acts as the supervisor and handles the access rights and the administrative roles, for each user on the system. It actually connect the address of the user with his rights. The administrator, can set and update for each user the access rights with regards to the following actions for each different ingredient:
Mint: with this access right, each user has the right to create a token which represents an ingredient
Transfer: allows user to transfer a token to other actors of the system
Receive: it allows user to receive a token after the transfer has been invoked.
Split: since a minted ingredient can have quantity on it, there are cases, where the user will have to transfer or use a portion of it, thus using split can create a new token with different quantity
Pack: lock the tokens that participate in the creation of a final product. For example a fruit salad, containing apples and strawberries, could be a final product, where apples and strawberries are the minted tokens.
The core of the system consists of two smart contracts. The former is responsible for the governance of the users, and the latter is responsible for maintaining all the information on the chain. The first contract acts as the supervisor and handles the access rights and the administrative roles for each user on the system. It actually connects the address of the user with his rights. The administrator can set and update each user’s access rights with regard to the following actions for each different ingredient:
Mint: with this access right, each user has the right to create a token that represents an ingredient
Transfer: allows the user to transfer a token to other actors of the system
Receive: it allows the user to receive a token after the transfer has been invoked.
Split: since a minted ingredient can have quantity on it, there are cases where the user will have to transfer or use a portion of it; thus, using split can create a new token with a different quantity
Pack: lock the tokens that participate in the creation of a final product. For example, a fruit salad containing apples and strawberries could be a final product, where apples and strawberries are the minted tokens.
The second contract implements the actions mentioned above and is responsible for tracking and maintaining all the data and the changes in the system. There are two primary data structures; the first describes all the properties of a token (ingredient), for example, the quantity, the owner, and the current holder; the second holds all the necessary attributes to maintain the information of the final food products. Furthermore, the implementation of this smart contract utilizes the functionality of mapping to keep the relations between the minted ingredients and the users who used to own them. Thus, the implementation of the functions required is the following. For each action, there is a private function to check the user access rights based on the management contract and contains the software logic required for this action. On the other hand, a public function is exposed to the blockchain network, which can be called and perform the proper actions by using the necessary private functions.
A token is minted by calling the mint token function, which takes two arguments, the ingredient ID and the quantity. Given that the account calling the function possesses the corresponding permissions, the token is created, while the auxiliary mapping that holds data about token ownership is updated.
A second action is the ability to exchange tokens between the users. In order to achieve this, the user has to call the transfer function, which will first check if the caller has the access rights to transfer a token, and if the receiver has the access rights to receive that token. If all the checks are validated, the transfer happens, and the aforementioned mapping is updated with the new changes. The final step in this action is the receiving process, which again checks if the user can receive a token, and if this is true, the mappings are therefore updated with the changes.
Another functionality required in the system is the ability to split tokens to create portions of them. Hence, the split function is responsible do that. It has two arguments, the token ID and the quantity. Again, there is the process of validating the user’s access rights and the ability to split the token based on the requested new quantity. The result of this function is the generation of a new token with the requested quantity, which is inserted in the mapping. The master token of subtracting the quantity is updated with the new remaining quantity.
The most crucial action in this contract is the pack function, which combines the requested tokens in order to create the final product. Once this function is called and all the required checks are valid, the tokens are consumed, and a new data structure is created. This new structure contains the tokens that compose the product and the owner, which is the company that produces the end product in most cases. All these data are stored in a new mapping, and there is a new hash generated to represent that product.
Finally, another function, the view pack, has view access rights and can be executed by anyone on the network; it offers the functionality of tracing all activity regarding the tokens that constitute the final product, along with the users that made the actions. This is mainly to be used by the end-user, which is the consumer that will purchase the food product. In this way, the system can provide him full traceability for the product in a secure and trustful way.
10. Conclusions
One of the most important issues facing consumers today is the origin of the foods and food-related products they consume, as a consequence of many food scares and the globalization of food markets, which has led to food mobility across countries and continents. Complexity and a lack of transparency define food supply networks. Customers’ demands highlight the crucial need for creative approaches for validating and certifying food’s origin, characteristics, and information. This paper discusses an end-to-end approach to secure food supply chain systems and contributes to meeting regulatory standards for transparency and traceability.
The system is based on a blockchain-based back-end that ensures data integrity, automation of workflows, and interoperability between monitoring systems of different vendors. A tokens-based scheme enables the detailed monitoring of food ingredients handling, transferring, or storing and allows for presenting such data for a specific food end product that reaches the consumer. Additionally, the prospect of developing a security system that utilizes TinyML’s fledgling technology to function as a safeguard for monitoring devices is studied. This approach is capable of identifying anomalies that occur when malicious actors try to exploit or tamper with devices that are destined to report data to the aforementioned system. It has been proven that it is feasible to embed an anomaly detection mechanism into monitoring devices of limited resources with satisfactory accuracy results.
To our knowledge, this is one of the first efforts, if not the first, that integrates two emerging technologies, blockchain and TinyML, with the goal of increasing food safety across the food supply chain. This work contributes to two areas: the transparency and traceability of food goods, which are increasingly requested by customers today, and security, as the anomaly detection device works as a deterrent to malicious actors.
Regarding future work, it is expected to test the tokens scheme developed in a broader set of supply chain applications, apart from the agricultural food industry, to revise or extend it accordingly to serve the needs of other domains as well. As the data collected may be of high importance, supply chain actors may be reluctant to share those with the public. A more versatile approach to providing privacy protection features is going to be developed on top of the current scheme through which users will be able to control the access to the data they store in the system. We also plan to apply the TinyML anomaly detection feature to a real-world scenario, in which actors of the supply chain will attempt to use the monitoring devices in their favor to validate that the methodology is appropriate.
Currently, a proof of concept implementation of the presented system has been developed in a laboratory environment. The next step is to apply the proposed system in a more realistic setup in collaboration with the food industry. This will enable us to validate the functional requirements for the system better and also assess any performance issues that may come up in order to fine-tune the system.
The final phase is to evaluate the system through user studies conducted with invited actors. Some of the critical points that need to be extracted from those studies are the users’ perceptions of the system’s interface and overall usability, their thoughts on how the system improved key components of the food supply chain, such as traceability, transparency, and sustainability, their perspectives of the mobile application, and finally, their overall thoughts and propositions for new system additions.
The present paper has introduced a novel approach in order to ensure end-to-end security for food supply chain monitoring systems. It presents the combination of blockchain technology and TinyML to enhance food supply chain security. While we have presented the concept, the design, and a proof of concept implementation, there is much room for the further development and validation of the proposed system.