V.RDA.2223 - Trakka - Making Data Flow
Exploring a producer centric data sharing infrastructure for the red meat industry.
Project start date: | 29 March 2021 |
Project end date: | 29 June 2023 |
Publication date: | 06 November 2023 |
Project status: | Completed |
Livestock species: | Grain-fed Cattle, Grass-fed Cattle, Sheep, Goat, Lamb |
Relevant regions: | National |
Download Report
(2.3 MB)
|
Summary
The Australian red meat industry has a wide range of technology solutions to improve business operational efficiencies, however the data produced by these systems remains frustratingly siloed, fragmented and inconsistent. Interconnecting these service providers and standardising the data shared between them could unlock latent value throughout the agrifood value chain. Any data sharing infrastructure must put data owners (producers) in control of their own data. The Trakka project explored the requirements of a producer centric data sharing infrastructure that could act as an ‘honest broker’ in the exchange of producer’s agribusiness data.
Objectives
1. Development of a self-sustaining data sharing architecture which enables stakeholders to share data throughout the entire value/supply chain within the red meat industry.
2. Development of Microservices using user input but focussed on three areas of business operations; compliance services, production efficiency services and new value creation.
3. Publication of guidelines, tutorials and development tools to enable service providers to utilise the event-based messaging service.
Key findings
A successful producer centric data exchange must:
- Put producers in control of their data
- Standardise all data transferred through it
- Allow any service provider to connect to any other service provider.
● A successful data exchange must be able to connect any data publisher with any data subscriber, on behalf of a producer. Configuring the architecture’s routing protocol to accommodate the vast number of potential connections represented a challenge.
● Data standards are foundational to any successful data exchange as they ensure everyone in the ecosystem is talking the same language. Data standards enable farmers to understand and accept new technologies and get more value from their digital assets. Data standardisation improves the value proposition and minimises the barriers to data exchange and reuse.
● The development of data publisher streams for both autonomous machine to machine connections as well as direct manually configured connections future proofs the technology. It also enabled the project to explore a number of architecture configurations.
● Evaluating the requirements for data subscribers to access data was tested using the data endpoint applications. The project tested both push and pull data configurations. It was recognised that service providers are currently familiar with a pull configuration, but in time push connections will become more commonplace and so accommodating both was important.
● Muzzle prints are like fingerprints that are unique to individual cattle. This project explored opportunities to use vision recognition software based on machine learning algorithms that could provide automated identification. This work demonstrated that there was potential for automated vision recognition software.
Benefits to industry
The main beneficiary for this project will be primary producers in the red meat industry. It provides the opportunity for producers to take control of their data and to more easily connect to new service providers. The underpinning technology provides opportunities for new value offering throughout the supply chain. It also provides the ability for new service providers to more easily access critical data permissioned by the producer that could result in unrealised value.
Future research
Further work needs to be conducted to manage the maintenance and extension of the common animal data schemas within an operational environment, including feedback mechanisms that provide producers and organisations the ability to comment on data standards.
Extending the muzzle recognition research to refine the image specification required is recommended to inform the hardware requirements before testing the technology in commercial supply chain trials can be conducted. The learnings from the further research and trials are required to deliver an accurate muzzle recognition system that can be pragmatically applied by Australian producers.
More information
Project manager: | Verity Suttor |
Contact email: | reports@mla.com.au |