ATLAS – Agricultural Interoperability and Analysis System
Modern farming is switching to a data-driven approach. A multitude of digital solutions, e.g., software systems, sensors, or other agricultural equipment is already available and used by farmers, and farmers have good reasons to use a specific tool. However, the interconnection of multiple tools, i.e., the exchange of data between these tools, can be very cumbersome due to the lack of interoperability. Connectivity and interoperability of systems and equipment used in agriculture are crucial factors in the digital transformation of the sector. The adaption of the most advanced digital technology in agriculture is the key for a more sustainable and productive food production, ensuring the supply of food and simultaneously reducing the environmental impact.
ATLAS Interoperability Network
ATLAS is a European research project funded under Horizon 2020, with 30 partners from 7 different European countries. The solutions developed in ATLAS will provide the means to interconnect agricultural software systems, agricultural machines, sensor systems and data analysis tools to establish the flow of data between them. ATLAS follows the approach of making well established and specialized tools interoperable, providing the possibility to decide which tool to use to the end-user.
The development of the ATLAS Interoperability Network is a bottom-up initiative involving stakeholders from all important agricultural domains, focusing on a technical interoperability which enables the exchange of information between different existing systems. As each of these participating systems is independent and built upon its own technical infrastructure, this leads to a distributed, non-centralized network of systems. In a nutshell, the ATLAS Interoperability Network will provide the means to interconnect different agricultural software systems and equipment and to establish the flow of data between them.
The basic building blocks to realize this flow of data are the ATLAS Services offered by the participating systems. These standardized services enable the exchange of data in a formally defined and documented way. The exchange of data between services is designed to be peer to peer without any network centralistic component for specific distribution or steering of data flows. This approach is mainly chosen by the fact that a central data forwarding component would be in the need to be capable of shifting an immense amount of data between the peers.
Although being a decentralized network of independent systems, central components are needed to guarantee the quality and reliability of the network. The main central component for this is the ATLAS Service Registry. It implements functionalityies for service discovery and for the registration of certified network participants. The Service Registry and its governance body ensure that any new service is conformant to the standard and that it meets quality-, safety and security requirements. This leads to a system of systems which is
- open yet regulated
- extendible but secure
- easy to implement but flexible
Furthermore, ATLAS works on a technical solution to bring interoperability to the edge of the Network, called the ATLAS AppEngine. The AppEngine is a reference specification for an ATLAS edge computing platform on which fit-for-purpose apps may be remotely installed. The AppEngine aims to provide a platform that supports operations in environments with little or no internet connectivity, such as for tractors operating in remote rural areas, but also for support scenarios that require very low latency when processing nearby sensor data to actuate adjustments in near real-time on local devices.
Conclusion
ATLAS develops solutions to achieve the goal of interoperability between sensors, agricultural machines, and data platforms. The architecture was designed with concrete use cases in mind. In this respect, two basic concepts are foreseen: data platform-based data exchange and processing, and edge-computing capabilities. With these complementary solutions, the implementation of complex use cases with real-time requirements and challenging in-field conditions will be possible, and a significant contribution to the digitalization of agriculture will be made.
