Tree-to-Me: Standards-Driven Traceability for Farm-Level Visibility

Abstract

Traditional horticultural information systems lack fine-grained, transparent on-farm event traceability, often providing only high-level post-harvest summaries. These systems also fail to standardise and integrate diverse data sources, ensure data privacy, and scale effectively to meet the demands of modern agriculture. Concurrently, rising requirements for global environmental, social, and governance (ESG) compliance, notably Scope 3 emissions reporting, are driving the need for farm-level visibility. To address these gaps, this study proposes a novel traceability framework tailored to horticulture, leveraging global data standards. The system captures key on-farm events (e.g., irrigation, harvesting, and chemical applications) at varied resolutions, using decentralised identification, secure data-sharing protocols, and farmer-controlled access. Built on a progressive Web application with microservice-enabled cloud infrastructure, the platform integrates dynamic APIs and digital links to connect on-farm operations and external supply chains, resolving farm-level data bottlenecks. Initial testing on Victorian farms demonstrates its scalability potential. Pilot studies further validate its on-farm interoperability and support for sustainability claims through digitally verifiable credentials for an international horticultural export case study. The system also provides a tested baseline for integrating data to and from emerging technologies, such as farm robotics and digital twins, with potential for broader application across agricultural commodities.

1. Introduction

In horticulture, growing demands for transparency and traceability are driven by the need to ensure food safety, sustainability, and environmental accountability. Consumers and regulatory bodies increasingly prioritise these practices to meet stringent environmental, social, and governance (ESG) standards [1].

Meeting sustainability goals relies not only on end-to-end supply chain traceability but also on active farmer participation via digital systems. This enables accurate Scope 3 emissions reporting, which refers to indirect emissions within the value chain that are not directly produced by the reporting entity. Agriculture significantly contributes to these indirect emissions, accounting for a substantial share of a retailer’s carbon footprint. Recent reports from Australia’s leading grocery organisations [2] highlight that addressing these emissions is vital for achieving net-zero targets. This underscores the need for robust systems that can accurately track and report key on-farm activities. Achieving long-term sustainability goals requires both the integration of diverse technologies and close collaboration among stakeholders across the value chain within a unified traceability framework [3].

The lack of interoperability between different systems and platforms offered by technology providers hampers progress in achieving full on-farm and supply chain traceability. The issue is more persistent at the farm level, where the diversity of data sources and formats hinders collective oversight. Seamless data exchange and integration of information generated or collected by these systems are critical for effective traceability, yet they are rarely enabled. The widespread heterogeneity in data models and semantics used to represent agricultural data, coupled with the absence of dominant standards and robust mechanisms to integrate agri-food information systems, further intensifies the challenge [4].

A potential solution is to apply global data standards, such as Global Standards 1 (GS1), to on-farm traceability. GS1 is a globally recognised system widely used in supply chain management. It incorporates unique identifiers, including the Global Trade Item Number (GTIN) [5], which identifies trade items such as products, and the Global Location Number (GLN), which specifies locations such as farms or blocks [6]. These identifiers enable precise tracking from farm to consumer, enhancing data consistency, interoperability, and sharing. However, GS1 adoption at the farm level remains limited. Using globally consistent identifiers under GS1 improves information management and simplifies data sharing across platforms and stakeholders. Moreover, aligning farm-level data with global standards like GS1 is essential for ensuring compatibility with downstream supply chain processes and supporting ESG reporting.

This paper introduces a farm event-driven traceability solution, Tree-to-Me (https://www.tree2me.com.au/, accessed on 20 April 2025), designed to systematically capture and integrate activities at the individual tree level using global data standards. The system bridges the gap between farm-level activities and supply chain standards, offering a practical and scalable solution to improve transparency and efficiency in agricultural traceability. Tree-to-Me is positioned as an example for industry adoption. With orchard-specific settings and event-based data integration, it facilitates the development of a digital twin for the orchard [7]. The system has undergone several case studies to validate its potential for real-world application and its technological robustness in supporting emerging data demands.

2. Related Work

Farmers increasingly rely on data from diverse technical providers and platforms in modern orchards to optimise practices and improve productivity. A critical requirement for such systems is interoperability: the seamless exchange and use of data across technologies. However, this is hindered by three interconnected issues: proprietary identifiers, limited or no support for data integration, and a lack of data standardisation.

Proprietary identifiers and inconsistent data standards imposed by solution providers hinder integration and interpretation. Internet of Things (IoT) sensors and farm management software each use unique data formats. Integrating these datasets requires their mapping to geospatial coordinates, standardising of naming conventions, and reformatting to ensure compatibility. This process demands extensive communication and introduces risks of human error. Without standardisation, systems cannot communicate effectively, and heterogeneity continues to hinder data sharing across acquisition and processing systems.

Various initiatives have emerged to tackle these interoperability concerns. ADAPT [8] is an open-source toolkit that simplifies collaboration between hardware and software solutions in precision agriculture by providing an agricultural application data model. Proprietary data models are mapped to a common standard, enabling precise compatibility and usability in specific agricultural contexts. Alternatively, the Open Agricultural Data Alliance (OADA) framework [9] takes a more flexible approach. It defines how software should interact with data through publicly available APIs and libraries. By removing strict pre-defined standards, OADA allows developers to build tools that adapt to varied agricultural contexts while ensuring secure and consistent data exchange.

Several models have also been developed to manage domain-specific data heterogeneity by adopting tuple-based metadata abstraction. The Heterogeneous Remote Sensing Observation (RSO) metamodel [10] proposes a five-tuple MOF-based structure to integrate remote sensing data. The SVASM model [11] manages real-time IoT streaming data by separating metadata into static and dynamic layers, extending the OGC observations and measurements framework. The HKPM framework [12] introduces a nine-tuple model for the representation of environmental KPIs in smart city contexts.

While these frameworks offer progress, most focus on local interoperability rather than aligning farm data with downstream supply chain requirements. Broader data sharing across the supply chain is essential for transparency and in supporting sustainability initiatives. Advanced traceability systems must increasingly align with global data standards to meet these evolving needs.

2.1. Global Data Standards for On-Farm Traceability

GS1 standards are widely adopted in supply chain systems to promote transparency, accountability, and interoperability. For example, food products sold in supermarkets in Australia and internationally carry barcodes that reveal key product information. By scanning these barcodes, consumers can access product details linked to the GTIN, a unique and globally recognised identifier. This seamless system allows products to be traced throughout the supply chain, enabling stakeholders to monitor processes, maintain product quality, and verify regulatory compliance [13]. The GTIN, along with other GS1 identifiers—such as the GLN for precise location tracking and the Serial Shipping Container Code (SSCC) for logistic units—forms the backbone of identification in traceability systems.

When combined with the Electronic Product Code Information Services (EPCIS) standard [14], which defines a framework for capturing and sharing visibility event information across supply chains, GS1 standards facilitate event-driven data sharing by capturing the “what, when, where, and why” of traceability events. These standards have demonstrated their utility in improving supply chain efficiency, ensuring consistent data exchange, and supporting seamless collaboration among stakeholders.

Research has highlighted the potential of integrating GS1 into agricultural traceability systems [15,16]. As a global standard, GS1 eliminates ownership restrictions and ensures universal recognition and validation across systems. Its unique identifiers simplify integration, allowing different technical providers to align their data with GS1 identifiers. GS1 also enhances interoperability by serving as a bridge between different systems and enabling seamless communication through its standardised framework. The proposed use of GS1 identifiers in the system, including GLNs, aims to enhance data interoperability and scalability by tracking farm locations [15]. Additionally, GS1 data standards facilitate efficient downstream data sharing and are intuitively applicable to farm activity traceability. Extending these standards to farm-level activities provides an opportunity to align production and post-farming data into a single service accessible to end users, including consumers and regulators. This integration has the potential to enhance trust, improve brand loyalty for farmers, and increase operational efficiency [15].

Despite these proposed designs and the clear benefits of GS1 standards, their adoption in agriculture remains underutilised, particularly in addressing the requirements of farm-level events. The Tree-to-Me system seeks to fill this gap by demonstrating the practical applicability of these standards in capturing farm activities while ensuring alignment with broader supply chain requirements.

2.2. Requirements for Data Transparency and Privacy

Apart from integrating data with global standards for sharing across the supply chain, another critical aspect is ensuring data authenticity. Globally, the European Union’s forthcoming Digital Product Passport (DPP) [17] is designed to uniquely identify products and will require a detailed record of each product’s lifecycle, including unique identifiers, compliance documents, and information on substances of concern.

A similar approach is being explored in Australia under the Australian Agriculture Traceability Protocol (AATP) [18], which extends such concepts to on-farm activities to deliver comprehensive product information aligned with evolving consumer and regulatory demands. The AATP builds on the United Nations Transparency Protocol (UNTP) [19], an international standard for traceability and transparency in value chains. Aimed at being industry- and geography-agnostic, the AATP offers a suite of foundational tools that enable the extension of UNTP to meet specific industry and geographical needs. For instance, the AATP has extended the UNTP-defined DPP to create a Digital Livestock Passport for Australian livestock characteristics and treatments. AATP is one of several UNTP extensions, with at least 100 expected to be registered in the coming years. The UN aims to issue one million product passports daily by 2030, supporting sustainability through credible claims about farm processes and resource use.

Emerging standards from the World Wide Web Consortium (W3C), such as Verifiable Credentials (VCs) and Decentralised Identifiers (DIDs), define methods for secure, tamper-evident, privacy-respecting, and machine-verifiable data exchange. VCs are digital statements issued by trusted authorities, containing information like licences, credentials, and identification. DIDs are decentralised identifiers that verify digital identities without relying on central registries. When linked, a VC and DID allow the holder to share provably authentic information, which verifiers can confirm without contacting the issuer [20]. However, the self-sovereign nature of VCs and DIDs means interoperability is not guaranteed, even between W3C-compliant tools. A unified protocol tailored to Australian agriculture is recommended to ensure consistency and effective implementation [21].

As data sharing increases, protecting privacy and ensuring farmer control are crucial. The Australian Farm Data Code (AFDC) [22] sets out principles to ensure farmers retain control over how their data are used and shared. Therefore, traceability systems must incorporate these governance principles to guarantee secure and responsible data handling.

3. Methodology and Implementation

3.1. System Architecture

The Tree-to-Me system is built on a technological framework, including a progressive Web app and a cloud-based microservice architecture, ensuring scalability, modularity, and robustness. The architecture is illustrated in Figure 1.

The adoption of a cloud-based microservice architecture is driven by the need for scalability, fault tolerance, and agility in managing heterogeneous agricultural traceability data. Microservices decompose systems into loosely coupled, independently deployable services. When deployed in the cloud, these services enable dynamic horizontal scaling. This allows resources to adjust based on demand, minimising downtime and enhancing system resilience [23,24]. The architecture also isolates faults within individual services, reducing system-wide failures and optimising cost efficiency by allocating resources only where needed [24].

The front-end system is a progressive Web application (PWA) designed to accommodate multiple user roles, including system administrators, farm owners, and farm staff. It supports access across a variety of devices, such as mobile phones, wearable devices, warehouse handhelds, and desktops. Built using modern Web frameworks, the application delivers dynamic, responsive interfaces with a native app-like experience. To support farms with limited or no connectivity, PWA capabilities are designed to function offline.

The application uses the Leaflet library [25] and Google Maps services [26] for spatial functionality, allowing users to create boundaries for farms and individual features. This spatial functionality not only provides a clear visual reference for users and third parties but also enables these areas to be assigned global identifiers. These identifiers allow farm data or specific features to be shared via digital links. The front-end Web application is deployed using a cloud-based hosting platform, ensuring reliable performance and ease of maintenance. Figure 2 illustrates the example Tree-to-Me Web interface for Sundial Orchard, located at Tatura Smart Farm, Victoria.

The back-end infrastructure leverages cloud services to manage both spatial and relational data. It uses a combination of geospatial databases and structured APIs to support data flow between the front-end and back-end systems. These APIs enable operations such as event recording, farm registration, and controlled data sharing with third parties via digital links. Access is managed through API keys, which are automatically generated and provided via the farm owner interface. The APIs are managed through cloud services to streamline requests and responses. These requests trigger automated extract–transform–load (ETL) operations in a serverless environment, ensuring on-demand resource allocation.

3.2. Data Standardisation

The Tree-to-Me system integrates global standards for data identification, data capture, and data sharing. During the sign-up process, farm owners must provide a GS1 company prefix, a 7–10 digit number assigned to their business, to be able to identify the farm in a globally unique manner. If a farm does not yet have one, Tree-to-Me facilitates the creation of a GS1 company profile, including the allocation of a GS1 company prefix. While the farm feature is registered, Tree-to-Me ensures the auto-generation and allocation of a GLN and its corresponding GLN extension. GLNs serve as unique identifiers for farm locations, while GLN extensions distinguish specific features, such as orchards, blocks, and trees. These identifiers create a parent–child relationship within the farm structure, ensuring seamless organisation. Each GLN extension is a serialised alphanumeric value of up to 20 characters, maintaining uniqueness and enabling precise tracking of resources.

When fruit is harvested and prepared for distribution, harvest shipments are identified using SSCCs. Each SSCC follows a structured 17-digit format comprising an extension digit (0–9) for scalability, a GS1 company prefix to identify the company, a serial number (six to nine digits) linked to the prefix, and a check digit for validation [27]. The check digit is calculated using a standardised algorithm: each digit in the 17-digit SSCC undergoes alternating multiplications by 3 and 1; the resulting products are summed, and the sum is subtracted from the next highest multiple of 10 [28]. This ensures numerical integrity, and SSCC accuracy is verified with an automated validation tool [29].

Currently, the Tree-to-Me system focuses on five critical on-farm events as an example of standardised data capture and sharing in heterogeneous data environments, enhancing data quality [30]. Figure 3 illustrates Tree-to-Me data collection guidelines, streamlining the management of farms, farm features (e.g., trees, rows, and blocks), and harvest-originated shipments.

This workflow supports traceability systems by capturing and managing two primary types of data: static master data and dynamic key data elements (KDEs) for critical traceability events (CTEs).

Master data include relatively static attributes defining farms (

Table 1. Master data for a farm considering sustainability-related brand development goals.

Farm
Attributes	Required	Data Type
Trading Name	Yes	String
Address	Yes	String
Key Personnel	Yes	UUID
GS1 Company Prefix	Yes	Char(7–10)
GLN	Yes	Char(13)
Boundary	Yes	Spatial Object
Certification	Optional	Multi-select List [ACOS, National Standard, EU, USDA NOP, COSMOS, IFOAM, Korean Cert, Japan JAS, Chinese Cert, Freshcare, HARPS, Non-GMO]
Farm Type	Yes	List [conventional, organic, biodynamic]
Farm Website	Optional	URL
Biodiversity	Optional	Multi-select List [habitat restoration, wildlife corridors, native vegetation, none]
Farm Energy	Optional	List [Solar, Wind, Fossil, Hydro]
Waste Management	Optional	Multi-select List [reduction, recycling, composting, none]
Regenerative Practices	Optional	Multi-select List [Cover cropping, crop rotation, no-till farming, agroforestry, none]
Soil Type	Optional	List [Red Earth, Black Earth, Sandy, Clay, Loamy]
Organic Carbon	Optional	Numeric (%C)
Soil pH	Optional	Numeric [0–14]
CEC	Optional	Numeric (meq/Kg)
Contact Email	Yes	Email
Contact Phone	Optional	Char(10)
Summary	Yes	String

), farm features (Table S1), and harvest-originated shipments (Table S2). These data must be provided when the object is created. KDEs for CTEs represent dynamic event data that occur throughout the production process, including activities like irrigation, chemical applications, pest control, fruit quality assessments, and harvest operations. KDEs tied to specific CTEs follow local and global jurisdictional record-keeping requirements such as the US FDA’s New Era of Smarter Food Safety Blueprint and Section 204(d) of the FDA Food Safety Modernization Act (FSMA) [31]. The collected event data align with the EPCIS standard, reflecting the five core dimensions—what, when, where, why, and how—that classify them as visibility events. For chemical/nutrient applications (

Table 2. KDEs for chemical/nutrient application following record-keeping requirements.

Chemical/Nutrient Application
KDE	Required	Measurement Unit	Data Type
Date of Event	Yes	N/A	Datetime
Farm Feature	Yes	N/A	Char
Product Trade Name	Yes	N/A	String
Type	Yes	N/A	List [Organic, Synthetic, Natural Mineral, Biofertiliser, Sustainable]
Dilution Rate	Yes	mL/L or mg/Kg	Numeric
Application Rate	Yes	L/Ha or Kg/Ha	Numeric
Application Method	Yes	N/A	List [Spraying, Dusting, Drenching, Fogging, Injection, Fertigation]
Reason for Application	Yes	N/A	String
Cover Crop	Optional	N/A	String
Wind Speed	Optional	km/h	Numeric
Wind Direction	Optional	N/A	List [N, NE, E, SE, S, SW, W, NW]
Temperature	Optional	°C	Numeric
Batch/Lot No	Optional	N/A	Char
Global Trade Item Number (GTIN) of Product Used	Optional	N/A	Char
Product Production Date	Optional	N/A	Datetime
Applicator	Yes	N/A	Char
Notes	Optional	N/A	String

), KDEs include the event date (when), the associated farm feature (where), the event type (what), the reason for application (why), and the application rate and method (how). Additional jurisdictional requirements ensure compliance with regulations, such as those governing agricultural and veterinary chemical use in Victoria [32]. Similarly, other critical events—irrigation (Table S3); pest, disease, and weed management (Table S4); fruit quality assessments (Table S5), and harvest (Table S6)—follow the same KDE definition process. Based on this workflow, which uses structured master data and dynamic KDEs within the EPCIS framework, the system can be extended to capture additional event types, supporting future traceability needs.

3.3. Data Management

The database design of the Tree-to-Me system is structured to implement the data ecosystem discussed above, providing fine-grained data associations and access/functionality control management. The underlying design is modular and scalable to accommodate future data capture needs, beyond the five example event types currently implemented in Tree-to-Me. Figure 4 represents the overall entity boundary diagram and database design of the system supporting this structure. Attributes indicated by an asterisk (*) represent primary keys.

3.3.1. Object Management

Object management is at the core of the database, encompassing the three objects related to farm operations and allowing for the collection and storage of the master data attributes for each object type. The three types of objects are managed via the object type table, which helps differentiate between the three object types and defines the master data attribute details specific to the respective object. This type of definition is then used by the object table, which manages all objects of all kinds created in the Tree-to-Me system by utilising the object type definition from the object type table. Furthermore, separate tables for each type of object are created to manage the individual master data attribute value for the given object. This method of data management simplifies and decouples data management for objects and can be easily extended to incorporate additional object types without interfering with existing ones. Additionally, it also gives a one-point definition of what data need to be collected for an object without having to change any other tables, enabling easy updates to align with emerging industry needs for master data management.

3.3.2. User Management

User management in the Tree-to-Me system is designed to enable role-based permissions for data and functions in the system. To date, three types of roles, namely admin, farm owner, and farm staff, have been created to simplify the whole farm data management process. The user table stores unique identifiers for each user auto-allocated when signed up or invited, as well as names, emails, and user types (e.g., admin, owner, and staff). This data management procedure is then used to control the permissions:

(1)

The admin role can approve or decline farm registrations by validating that the farm has provided adequate and valid details. Admins can also disable or delete a farm. They can invite other admins to support change management and serve as the primary point of contact for system communications. Automated system emails are sent on their behalf as part of sign-up processes.

(2)

The farm owner’s role may be as the owner, manager, or key personnel of the farm responsible for farm creation. The farm owner can edit master farm data, create farm features and master feature data, create shipments and master shipment data, and create or edit any event and its key data elements.

(3)

The role of farm staff is allocated by invitation only by the farm owner, who may wish to onboard staff to work on the farm. Farm staff can perform all tasks the farm owner can, except for farm creation and the editing of master farm data, in addition to being unable to invite other farm staff themselves.

Role-based permission control is managed by the overall Tree-to-Me system as part of different components and services within the system; therefore, it does not need to be explicitly managed at a database level.

3.3.3. Event Management

The Tree-to-Me system is designed to capture and organise data related to various farm activities that can occur on the farm. Numerous critical traceability events may exist on a farm, for example, five CTEs were identified as visibility events that may be of interest for traceability purposes, allowing growers to build a brand around their sustainability and eco-credential practices. The selected events are irrigation, harvest, pest/disease/weed assessments, fruit quality assessment, and chemical/nutrient application.

In order to simplify data management for events, an event type table is created, as shown in Figure 4, that allows different event types to be managed with one-point references for key data elements that can then be used to collect the values and measurements required for the event. This allows an extendable way of managing data for events, adding new event types and respective key data elements and an easy-to-update key data element list for existing events as required without affecting any existing data. Such an approach is desirable for traceability, as the requirements for data collection and management may change over time given regulatory, market access, or consumer behaviour changes.

Additionally, the measurement units and range of data elements may differ across jurisdictions and markets for which the Tree-to-Me system can be suitably fine-tuned. Each event is linked to an object that the event was carried out on, as well as the user executing it on a specific date and time; the remainder of the data needing to be collected vary between events.

3.3.4. Access Management

Access management ensures that data are accessible to authorised users while maintaining security. GS1 digital links are used for encoding of the object identifiers, which are then used for data sharing. Figure 4 shows access management tables that help store digital links for traceability by including only the global traceability identifiers that can be understood by other systems.

The UUIDs are not shareable beyond the system, which may prohibit decentralised data sharing. Digital links are auto-generated by the Tree-to-Me system in the form of QR codes, along with shipment labels following GS1-128 barcode standards. The data management procedurefor this allows for fine-resolution control over what data can be shared. At harvest, each generated shipment is linked with the parent farm feature from which the shipment originated. At this point, growers can choose all or individual farm/feature/shipment attributes and event KDEs to be shared via the digital link. This approach of grower-driven control allows for effective privacy and security management such that the configuration allows growers to only share data required for the targeted supply chain receiver, without worrying about sensitive data access if they do not want it.

Furthermore, any third-party access via the digital link can be controlled for a date range for the events on the farm, as well as a date range for data to be accessible via the digital link. All digital shipment links only have a maximum of 12 months of access from a given date, as the shipments are not expected to be in the supply chain beyond that time line according to industry standards. This allows the grower to share targeted data for different markets for their brand development as suited. Additionally, all third-party data access is timestamped and recorded for security purposes, providing a reference point for access management.

3.4. Data Exchange and Sharing

Data exchange and sharing in the Tree-to-Me system are employed with two primary mechanisms: (1) dynamic API-based data exchange for on-farm data, enabling controlled access to farm event records through public APIs, and (2) digital link-based data sharing for both on-farm and shipment-related information, facilitating seamless tracking and management of harvested produce extended to the supply chain.

3.4.1. API-Based Data Exchange

The Tree-to-Me system provides API endpoints that enable public access to farm event data while maintaining data security and integrity. Access to these APIs is managed using API keys, which are automatically allocated and shared with authorised users through the farm owner interface. The farm owner defines the type of GET or POST permissions granted to specific API consumers and selects the list of endpoints accessible to them. The APIs are deployed and managed using the Amazon API Gateway, which ensures efficient request handling and response delivery. AWS Lambda functions are used to perform extract–transform–load (ETL) operations on the database in a serverless environment. This design enhances flexibility and scalability while allowing for modification or extension of API services without disrupting existing workflows.

To access the Tree-to-Me public farm event data API, users must be invited by a farm owner. Once invited, the following steps are required to authenticate and retrieve data:

(1)

Receive invitation email: The Tree-to-Me system sends an email containing the API key ID, API key secret, and a link to the API documentation.

(2)

Obtain JWT access token: The user calls the OAuth token API using the provided API key ID and API key secret to request a JWT access token.

(3)

Query the farm event data API: The JWT access token is used to query the farm event endpoint to retrieve relevant farm event data. A validation procedure is put in place for data access to ensure API consumers can only access permissioned data for a given farm.

(4)

Refresh expired access token: If the access token expires, the user must call the OAuth token API again with the key credentials to obtain a new access token.

Table 3. Tree-to-Me Public API.

Tree-to-Me Public API
Method	Endpoint	Description	Query/Body Parameters
GET	get_pest_disease_weed_events	Retrieve a list of pest, disease, and weed assessments Event Data based on the provided query parameters.	farm_gln, gln_extension, start_time, end_time optional: time_order, limit, offset
	get_fruit_quality_events	Retrieve a list of fruit quality assessments Event Data based on the provided query parameters.
	get_chemical_nutrient_events	Retrieve a list of chemical nutrient applications Event Data based on the provided query parameters.
	get_irrigation_events	Retrieve a list of irrigation applications Event Data based on the provided query parameters.
POST	oauth	Used by Public Users to exchange JWT access token to public data API by key id and key secret.	key_id, key_secret
	create_pest_disease_weed_event	Create a new pest, disease, and weed assessment with the provided data.	farm_gln, gln_extension, event_data (species, symptoms, control_measures, damage_assessment, notes)
	create_fruit_quality_event	Create a new fruit quality assessment with the provided data.	farm_gln, gln_extension, event_data (assessment_type, sweetness, firmness, defects, colour, notes)
	create_chemical_nutrient_event	Create a new chemical nutrient application with the provided data.	farm_gln, gln_extension, event_data (application_type, method, dilution_rate, application_rate, environmental_conditions, notes)
	create_irrigation_event	Create a new irrigation application with the provided data.	farm_gln, gln_extension, event_data (irrigation_type, water_source, application_rate, energy_source, notes)

illustrates available public APIs and their role in exchanging farm event data within the Tree-to-Me system. More details on the API are available at https://documenter.getpostman.com/view/33833398/2sA3XPCN7F#intro (accessed on 20 April 2025).

3.4.2. Digital Link for Decentralised Data Sharing

The Tree-to-Me system also employs GS1 digital link technology to facilitate data sharing. This approach ensures that relevant stakeholders can access farm and harvest event data, both at the farm level and throughout the supply chain.

The GS1 digital link is a standard method used to transform a product’s GS1 identifier into a Uniform Resource Identifier (URI), allowing users to access detailed product information. Tree-to-Me system encodes GS1 identifiers, including GLNs for farm features and SSCCs for harvested consignments, in URL format. These URIs are embedded in QR codes or barcodes or written to NFC tags. Users can scan these to access the linked online information directly. The GS1 digital link can also be decoded to obtain the global identifiers without requiring a centralised system connection. It can then be further validated using services such as Verified by GS1 to retrieve metadata, including farm business information associated with the identifiers.

First, the farm feature digital link is used to identify farm features and farm events associated with specific locations within the farm. This creates a parent-to-child relationship between the farm and its features. User roles with the appropriate permissions must be created to enable these updates. Once in place, farm owners and authorised users can insert or modify farm event data, ensuring that current information is maintained. An example QR code for a farm feature is shown in Figure 5, available via https://www.tree2me.com.au/414/9377779611722/254/z%3A%275JPN%3EU%27bZHD%2B%27zgoy (accessed on 24 February 2025). Second, the harvest digital link is used to share harvest event data as the produce moves through the supply chain. An example QR code is shown in Figure 6, available via https://www.tree2me.com.au/414/9377779611722/00/093611640000000017 (accessed on 24 February 2025). The GS1 128 barcode and QR code standards are used to encode the SSCC, ensuring compatibility with traceability systems used across the industry. Supply chain participants can access the permissioned data and export them for further use within their own systems.

The digital link encodes identifiers and is used to manage role-based access control for authentication and authorisation of data access. This follows the National Farmers’ Federation’s Farm Data Code and the Food Agility Cooperative Research Centre (CRC) Data Sharing Template. This design supports a decentralised data ecosystem, allowing data to be shared without the need for system-level integrations. It makes the Tree-to-Me system the first of its kind to adopt the Australian Agricultural Traceability Protocol for horticultural application in Australia.

3.5. Overall Workflow

Figure 7 presents a workflow diagram of the Tree-to-Me system. It outlines key processes and illustrates how different stakeholders interact with the system to capture, manage, and share data. The workflow demonstrates how the integration of global standards enables efficient and transparent data exchange.

At the initial stage, when a new farm is onboarded into the Tree-to-Me system, the farm owner registers the farm and submits their GS1 company prefix, which uniquely identifies the business within the GS1 network. The system administrator then reviews and approves the registration. If the farm does not already have a GS1 company profile, the administrator facilitates its creation. A GLN is then generated to uniquely identify the farm’s physical location.

Once registration is complete, the farm owner configures the farm’s digital structure within the system. This includes defining boundaries and adding geospatial features, such as orchards, blocks, rows, and individual trees.

With the farm structure established, data collection can begin. The farm owner invites farm staff to contribute event data using the system’s Web or mobile interface. Staff can record new farm features, such as adding new rows or individual trees, and log farm activities. For automated data collection, the farm owner grants API access to technology providers, allowing equipment such as sensors and IoT devices to transmit event data directly into the system. Farm visitors attending events can scan QR codes or NFC tags associated with specific farm features or locations to retrieve public master data or permissioned event data instantly.

As fruit moves from the farm into the supply chain, GS1 digital links and barcodes encapsulate key product information. These allow downstream supply chain participants to access permissioned event data related to specific consignments.

Throughout the entire process, farm owners retain full control over how their data are shared and accessed. When external users request access to the Tree-to-Me system’s public farm event data API, they must first be invited by a farm owner. Upon approval, they receive an API key ID, API key secret, and a link to the API documentation, ensuring that all data retrieval processes are properly authenticated and secure. Additionally, the Tree-to-Me Web interface provides customisable data sharing settings. This allows farm owners to manage the visibility of their data while promoting transparency across the agricultural value chain. An illustrative example is the export of a cherry consignment from Cantrill Orchard to China. A QR code embedded with a GS1 digital link provided access to the digital product passport (DPP) and Digitally Verifiable Credential (DVC), ensuring the accuracy and verification of export block details and associated event records.

3.6. On-Farm Interoperability Tests and Case Studies

The Tree-to-Me system was piloted on Agriculture Victoria’s SmartFarms for pome and stone fruits in Tatura and for almonds in Mildura, Victoria. The pilot involved defining spatial boundaries and features, capturing events through visual interfaces and automated sensors, and sharing data using GS1 standards. The results demonstrated the system’s ability to assign unique identifiers to farm features, track activities with precision, and share data seamlessly across platforms. In addition, several case studies were carried out with industry partners to validate the automated and systematic approach to farm event data capture. With the integration of automation technologies, the system also demonstrated the ability to insert farm events automatically, reducing recording workload, enabling real-time updates, and minimising human error compared to manual logging. All data were collected with farmer consent and managed in accordance with data governance policies.

3.6.1. Talgil Sensor Integration for Irrigation Events

Talgil [33] is a leading provider of precision irrigation control systems. It offers advanced sensors installed in water valves that accurately monitor water usage and allow for remote adjustment of irrigation settings.

Talgil’s integration into the Tree-to-Me system uses cloud services and APIs to automatically retrieve and process irrigation event data. The data are first collected from Talgil’s API, then standardised using Azure Cloud Services [34], which triggers periodic updates. Once standardised, the data are sent to the system through its public API. By automating this process, farmers benefit from a streamlined data flow, enabling efficient and accurate irrigation monitoring without manual recording.

For standardisation of data from the Talgil system, the extract–transform–load (ETL) function maps the proprietary ID used for water valves to the GLN extension in the Tree-to-Me system, corresponding to the associated irrigation block. The captured irrigation data follow system guidelines to extract the relevant key data elements (KDEs) for the irrigation event before being inserted into the system. While a specialised ETL process was developed to demonstrate the standardisation approach for farm-level interoperability, it is expected that solution providers will increasingly adopt global standards, eventually making explicit mappings unnecessary.

In the case of an irrigation event inserted for the Sundial Orchard in Victoria’s SmartFarms, eight blocks are monitored, and the Talgil system is integrated with Tree-to-Me and cloud service. Irrigation vents are recorded manually during irrigation but can be obtained in real time directly from the API.

3.6.2. Rubens Handheld Device Integration for Fruit Quality Assessment

Rubens Technologies [35] provides a handheld device with fluorescence-reflectance sensors integrated with machine learning analytics for rapid fruit quality assessment, replacing traditional destructive testing methods that are time-consuming and limited to small sample sizes.

The device connects to a mobile application via Bluetooth, enabling seamless data transfer to a cloud service through APIs. In the cloud, machine learning algorithms analyse the data to predict key fruit quality attributes, such as firmness (kgf) and sweetness (Brix), supporting informed management decisions aimed at improving orchard profitability and consumer acceptance. Growers benefit from rapid fruit quality evaluation without regional restrictions, as the portable device allows for easy, on-the-spot measurements.

The integration of the Rubens Technologies device into the Tree-to-Me system leverages cloud connectivity to automate the association of sensor-based fruit quality measurements with orchard features such as individual trees or fruit bins. To support this integration, the Rubens app was enhanced to read and decode GS1 digital links (QR codes) associated with farm features (trees, rows, or blocks) based on global standards. Following the existing Rubens Technologies process, the measured fruit quality data are automatically mapped to the KDEs required for the fruit quality assessment before being uploaded for the corresponding block by identifying globally unique IDs in a decentralised manner.

In this particular case, it is demonstrated that solution providers can undergo changes to be compliant with global standards and independently and comprehensively report farm events to a collective data ecosystem for the farm. This enables downstream supply chain integration and traceability. Unlike the Talgil use case, the Rubens system underwent standard adoption itself, thereby voiding any manual mapping requirements. This approach is encouraged for solution providers to achieve true on-farm interoperability and enable end-to-end traceability.

3.6.3. Digital Twin Integration for Sustainable Orchards in a Future Project

The Tree-to-Me system supports the consumption of even data in an example digital twin model for Agriculture Victoria’s Tatura SmartFarm as part of the Sustainable Orchard for a future project [36] led by Food Agility CRC in partnership with Agriculture Victoria, RMIT University, and the University of Queensland. This project aims to develop a data-driven digital twin that provides immersive 3D visualisations of physical orchard assets, combined with advanced analytics for interpretation of seasonal and operational datasets.

Through the system’s public API, designed for data sharing, the digital twin gains access to both real-time and historical irrigation logs (for example, integration) using the GET events API. When querying the API, decentralised global IDs such as the GLN/GLN extension for farm features are used. The KDEs and master data obtained by querying the system using the global identifiers then inform the digital twin for that corresponding feature.

By streamlining real-time and historical event data from orchard environments, the system enhances orchard digitalisation and sets a foundation for the development of the digital twin model. Through this case study, the Tree-to-Me system validated the use of a standards-based, secure data-sharing framework, enabling advanced analytics and visualisation while reinforcing data privacy and control for farmers in the development of emergent value-added productivity tools.

3.6.4. AgTrace Project—Horticulture Pilot

The Tree-to-Me system was presented to the Department of Agriculture, Fisheries, and Forestry (DAFF) as a horticultural traceability Proof-of-Concept (PoC) pilot as part of the Food Agility CRC and DAFF’s AgTrace solution under the AgTrace project undertaken by Food Agility CRC, adhering to the Australian Agriculture Traceability Protocol (AATP). It also enables the integration and exchange of orchard traceability data in the form of a digital product passport (DPP) with a supply chain system, using the scenario of exporting Australian-grown cherries to China.

The system facilitates orchard event data sharing using GS1 digital links for threefold usability, ensuring supply chain traceability, product origin verification, and sustainability claim verification. The data shared via digital link are consumed by FreshChain [37], a secure, blockchain-based paddock-to-plate assurance system that then issues the orchard DPP and digitally verifiable credential (DVC) for the orchard’s export registration with DAFF. Additionally, Source Certain International [38]’s Australian origin verification service issues the DVC that certifies the product’s origin, which is also integrated with the orchard’s DPP.

The use case involves cherry export consignment from Cantrill Orchards in Nashdale, NSW, to China. It showcases how supply chain partners could securely access farm event data and master data for DPP and DVC creation and ingestion into the supply chain. The trial successfully demonstrated high-resolution traceability for farm events and master data carried through the supply chain for complex export scenarios.

4. Discussion

The Tree-to-Me system is a pioneering solution applying global data standards to farm-level operations to enable orchard traceability. By leveraging global identifiers for data identification, capture, and sharing, it addresses the limitations of existing traceability systems and integration barriers, ensuring compatibility with broader supply chain processes. Specifically, the system aligns with the AATP and the UNTP, enabling compliance with national and international traceability guidelines while maintaining interoperability within a farm context that adheres to international frameworks.

Currently, the system records five core on-farm activities: irrigation, chemical application, pest control, fruit quality assessment, and harvest. Future implementations aim to expand this scope by identifying additional critical traceability events (CTEs) to support the recording of transaction, aggregation, and transformation events across the farm. Integrating the creation of digitally verifiable credentials within the Tree-to-Me system would further enhance the validation and reliability of the collected and shared data. While the initial focus is on pome and stone fruits, the framework is designed to be adaptable to other crops, commodities, and user groups, making it versatile for broader agricultural applications.

The Tree-to-Me system has undergone enhancements to adopt World Wide Web Consortium (W3C) standards [39], aiming to capture on-farm data in machine-readable formats for seamless sharing with supply chain partners. Although not fully implemented, this approach ensures that data can be exchanged across diverse platforms without losing context or accuracy. In parallel, the concept of digitally verifiable credentials adds another layer of integrity by offering tamper-proof certificates that can be attached to products at the farm level. These credentials, adhering to W3C recommendations, enable the secure verification of sustainability claims—such as efficient water usage or reduced chemical inputs—without risking unauthorised duplication or manipulation. By aligning with global standards, prioritising controlled data sharing, and incorporating verifiable credential frameworks, the Tree-to-Me system is poised to meet emerging sustainability and traceability requirements while maintaining robust interoperability and data privacy.

Farmer participation in digital traceability systems is critical for the achievement of sustainability targets, particularly as retailers pursue net-zero emission goals. In Australia, leading retail organisations are aligning with the Science-Based Targets Initiative (SBTi) to reduce emissions, with a focus on Scope 3 emissions, which refer to indirect emissions arising from production, processing, and distribution. Agriculture is a major contributor to these emissions, with the majority originating directly from the production process. For instance, Woolworths has enrolled 78 suppliers in its emissions programme, covering 25% of its Scope 3 emissions, to improve transparency and track progress. To meet sustainability criteria, retailers increasingly prioritise sourcing from farms that adopt eco-friendly practices. This incentivises farmers to use on-farm traceability systems that adopt GS1 standards, enabling secure management and sharing of farming activity data throughout the supply chain. These systems facilitate precise emissions tracking and empower stakeholders to identify reduction opportunities.

Despite these advancements, challenges persist. A significant amount of manual work is still required, limiting the system’s adaptability to evolving reporting requirements. For example, changes in reporting structures or data standards could disrupt the system’s functionality. Integrating AI agents into the Tree-to-Me system can further simplify on-farm traceability for growers and streamline compliance with emerging standards. By developing AI models that consume GS1 identifiers, W3C protocols, data standards, and sustainability criteria—including future changes—farmers can continuously validate whether their practices meet specific regulatory or market requirements. Instead of navigating lengthy technical documentation, growers could rely on an automated system that interprets, monitors, and flags compliance issues in real time. This approach would reduce manual data entry and oversight while also providing actionable insights to help farmers adapt their practices more effectively. The Tree-to-Me system is already being proposed for further enhancements and research to support these capabilities.

In this envisioned data ecosystem, the Tree-to-Me system would manage the collection, storage, and sharing of farm data in accordance with global standards. AI agents would interpret sustainability protocols and extract data from various sources, including CSV files, PDFs, and other unstructured formats [40]. By mapping these data to specific reporting requirements, the AI could generate a coherent output that integrates evidence from multiple sources. This capability would help growers and supply chain stakeholders maintain accurate, up to date records; reduce administrative overhead; and ensure consistent communication of sustainability metrics across diverse platforms.

The Tree-to-Me system ensures that different technologies can effectively communicate and operate through standardised frameworks. The system also benefits from the gathering, validation, and sharing of data to support sustainability initiatives. The Tree-to-Me system demonstrates significant potential to drive innovation in agricultural traceability and planning for W3C-compliant technologies and the exploration of AI-driven solutions.

5. Conclusions

The Tree-to-Me system marks a transformative advancement in horticultural traceability by integrating global standards into farm-level operations. Designed to track key on-farm events—such as irrigation, chemical applications, pest control, and harvesting—the system ensures interoperability, data security, and farmer-controlled access. Initial testing on Victorian farms and example solution providers across the value chain has have demonstrated its effectiveness in improving data accuracy, supporting compliance reporting and enabling sustainability claims through the use of digitally verifiable credentials.

The key advantage lies in its use of global standards to address data heterogeneity and proprietary format issues. It supports farm-level interoperability and connects seamlessly with the broader supply chain, unlike models limited to local integration. The system captures both static master data and dynamic key data elements (KDEs) linked to critical traceability events (CTEs), enabling flexible and standardised data flows. Its integration of cloud-based microservices and dynamic data sharing enhances scalability and supports emerging use cases with large suppliers seeking to meet sustainability initiatives, such as Scope 3 emissions reporting. The current baseline implementation of the system has already proven to be technologically robust and aligned with emerging standards. Nonetheless, challenges such as manual processes and evolving reporting requirements underscore the need for future advancements, including the adoption of World Wide Web Consortium (W3C) standards for machine-readable data and the incorporation of AI agents to automate data extraction and mapping.

By aligning farm-level traceability with global supply chain standards and addressing critical gaps in scalability, data privacy, and interoperability, the Tree-to-Me system demonstrates significant potential for broader adoption. Its foundational design sets the stage for further innovations—such as integration with robotic agents and the expansion of digital twins—ensuring its continued relevance in the evolving landscape of sustainable farming.