This section covers various aspects of our proposed solution for building a fruit and vegetables traceability platform. We start by presenting a use-case model that helps identify actors and their activities on the platform. After that, we discuss the domain model, highlighting the data we need to store and the methods we use to trace quality and sustainability indicators. Finally, we showcase the proposed architecture for this solution.
A background of fresh produce blockchain traceability: Identification of Actors and Their Activities
Based on the analysis of the results obtained in the previous section, we have identified three actors operating in the value chain. To illustrate the activities of these actors on the platform, we have created a use-case model depicted in Figure 4. The model provides a comprehensive overview of the roles and actions of each actor, enabling us to gain a deeper understanding of the underlying processes and interactions taking place on the platform.
Computers.
CHAIN-TRACE Use-case model for the fruit and vegetables traceability platform.
A background of fresh produce blockchain traceability: End Consumer
The Final Consumer represents anyone interested in the traceability of lots and information about them. This actor does not need to log in, being able to consult information about products and lots, as well as trace them back to their origin, along the lots’ transformations and mixings throughout the value chain.
5.1.2. Operator (Normal or Administrator)
This actor is more generic, representing two types of operators that can perform functions in this traceability platform for the fruit and vegetables value chain, after logging in. The first one, known as the Normal Operator, can perform all the actions that a Final Consumer can, plus registering additional activities such as harvesting, transport, and treatment. The second type, the Administrator Operator, has all the capabilities of the Final Consumer but also has additional privileges such as updating public information about the organization, details about measurements and products, and adding new products and activity types.
Fresh produce blockchain traceability Platform Administrator
This actor represents an entity that manages organizations in the value chain.
5.2. Data to Store for Tracking Quality and Sustainability
Once all stages and actors involved in the platform have been analyzed, it becomes crucial to determine the data model, or data structures, of the information to store for the platform to function correctly.
Our approach will utilize a “Blockchain + Database” methodology, which necessitates distinguishing between data stored on the blockchain (on-chain) and the database (off-chain). On-chain storage only includes data essential for maintaining traceability (such as lots and their associated activities). On the other hand, off-chain storage contains all other data (such as organization and product information). A visual representation of all the domain entities treated in our traceability platform is depicted in Figure 5 in the form of a domain model. A detailed overview of those domain entities is presented next.
Computers 13 00112 g005
Organization, Organization Type, Allowed Activity, Operator and Role—Organization represents an entity linked to activities carried out on our platform, defined by their name, phone number, email, and location (coordinates). Organization Type represents what type of organization it is (e.g., transport, control, and treatment), defining what activities it will carry out on the platform. Allowed Activity results from a m2m relationship between the Organization Type and Activity Type tables, allowing the definition of the types of activities that organizations can carry out on the platform. The Operator will represent an employee of an organization, being responsible for taking action on their behalf, defined by the code and password that allow him to log in to the platform, with a private code only recognizable by the organization where he works by his role. This will determine whether it is a Normal/Administrator Operator or a Platform Administrator. Another two fields (active and logged) will allow us to know whether the operator is logged in and whether a user is active or not, i.e., whether they will be able to log in and carry out activities on the platform. The role represents the different roles in the platform.
Product, Product Type and Lot—Product represents any product in our system (i.e., any fruit or vegetable, whether or not it’s transformed), defined by its name, description, and type. Lot represents a product in quantity and with any event performed on it. Product Type represents the different product types in the platform (e.g., fresh and transformed).
, Measurement, , Unit, —stores a type of activity (e.g., harvest, transport, and storage). Measurement saves the name of important and specific conditions for each activity (e.g., temperature, humidity, and fuel), stores an actual value to those fields (e.g., 45), and Unit saves the name and symbol of a measurement (e.g., degrees Celsius and the corresponding symbol “ºC”). Mandatory measurement results from the many-to-many relationship between the Measurement tables, allowing us to know the obligatory measurements by activity type.
Status—Status represents the current stage of an activity (e.g., started, ongoing, and finished).
InputLot and OutputLot—InputLot is the lots going through an activity, and is the lots that originated from that activity. Both tables result from a m2m connection between the activity and lot entities.
Activity—Activity represents the joining of all the entities described until now, defined by the operator in charge of the activity, the organization where the activity starts, the organization where the activity will end, the start and end timestamps, the type, the product, and the status.
Token—Token is going to store all the active refresh tokens.
A background of fresh produce blockchain traceability Tracking Quality and Sustainability
The effective traceability and management of sustainability and quality indicators, together with a product’s lots traceability, along a value chain, require the collection and storage of such indicators.
Several sustainability and quality indicators can be gathered, such as water, to enhance efficiency [43] and monitor quality [44]: waste, which can be employed to pinpoint sources of food waste in the value chain [45], and soil, which can ensure soil fertility is maintained [46]. It is vital to collect and store these indicators, as they can aid in improving food safety, mitigating food fraud, and reducing environmental and ecological impacts in the food production process.
The proposed platform allows the definition of the indicators to be tracked, including any metric associated with a production or logistics activity. As seen before in the domain model (see Figure 5), the modeled entities enable the definition of Mandatory Measurements/Measurements (e.g., temperature, humidity, and fuel) for each Activity Type (e.g., harvest, transport, and storage). And, each Activity of that Activity Type must have Taken Measurements for those Measurements. A Taken Measurement is an actual value of a Measurement. The platform is then able to record measurements for any defined measure associated to an activity type. In Table 2, examples of measurement indicators that can be collected and saved per activity are illustrated.
Table 2. Example of some quality and sustainability indicators per activity.
Any activity has mandatory measures to store, defined by the Mandatory Measurement entity. In addition to the mandatory measurements, organizations may want to save others. The Taken Measurement entity represents all the effectively taken measurements: in more detail, their value and timestamp. The Measurement entity represents the measurements (name and associated unit) that organizations wish to save. Unit stands for the unit’s name and the symbol that represents it. Figure 6 illustrates a small example of the process.
A background of fresh produce blockchain traceability Example of the storage process of quality and sustainability indicators.
Because the quality and sustainability indicators are not necessary to keep products’ lots traceability immutable, we used the database (off-chain). The three measurement tables, associated with the respective entities in the model, and the unit table are used to keep them stored.
5.3. Architecture
After analyzing all the previous information, we ended up with a three-tier architecture for the fruit and vegetables traceability platform, developed using the MERN Stack. The three-tier architecture comprises three distinct layers—presentation layer (frontend), application layer (backend), and database layer (storage)—each serving a specific purpose. We provide a simplified representation of the architecture