About our Project!

In recent years, agriculture has confronted numerous challenges, including the need to feed a growing global population, workforce shortages, sustainability requirements, and environmental constraints. The digitization of the agriculture industry faces various technical and socio-economic obstacles. Agricultural data, originating from diverse sources like individual farmlands, animal facilities, and IoT devices, poses challenges in terms of quality, security, storage, and the absence of decentralized data management systems, hindering the adoption of smart farming practices. Traditionally, this data has been processed in the Cloud, introducing limitations related to timeliness, privacy, localization, latency, and operational costs. To address these challenges, edge/fog computing has been employed, processing data closer to its source, coupled with low-power edge systems and efficient last-mile communications management to optimize energy consumption.

The AGRARIAN project aims to develop tailored open-source digital solutions, focusing on creating an open and dynamic environment (TRL 5-7). This environment will emphasize tools, reuse, composability, and orchestration, leveraging mature technologies for different workloads, including batch analytics, interactive, and streaming data. A new programming environment and tools will be developed for automated refactoring of applications and data management frameworks to operate in edge-based deployments. AGRARIAN will introduce edge-native programming abstractions and middleware, using widely accepted open-source technologies and container-based technologies like Kubernetes. This dynamic environment will accommodate various workloads, ensuring high responsiveness and efficient backhaul bandwidth utilization.

Major efforts in AGRARIAN will address challenges related to sharing data, models, and processing across AI clients and applications. This includes managing model accuracy, and fairness, and addressing different types of biases that may arise due to the contextualized use of AI. The project will leverage hardware-level, localized data analytics, and security-centric acceleration, and deploy strategies for detection, prevention, and containment at the edge. A robust communication infrastructure will support mobile and nomadic applications in areas with weak connectivity, reducing latency, increasing bandwidth, and offering a high degree of automation for quality-of-service performance.

The AGRARIAN platform aims to provide smart connectivity, dynamically meet the requirements of different network technologies, integrate edge and cloud transparently, ensure uniform management of distributed digital infrastructure, and optimize operations for energy consumption. It will include a robust Agricultural Decision Support System (ADSS) considering the specificities of farmers in rural communities. The ADSS will be data, communication, and knowledge-driven, with a simple GUI, re-planning components, and prediction abilities to adapt to dynamic situations.