In IoT-based agricultural systems, precise real-time decisions are crucial. Data synergy strengthens model efficacy, and to leverage agricultural data-driven models, data flow, both for inputs and outputs, is necessary. Data from sensors, vehicles, UAVs, and other sources can be made interoperable by leveraging event-driven architecture in data transactions among data actors, which enforces real-time decision support. To strengthen the data from the sources above, context and related situational information about data, which can be gathered from public repositories such as weather stations and soil survey databases, are essential. In most cases, these repositories provide data as a large file with a distinct data structure, potentially restricting interoperability. In this study, we propose a framework that uses the concept of event-driven structure to parse and distribute the data from public repositories on demand. Using an XML parser, this framework parses public weather data (present condition and forecast) using a weather API and soil properties from the SSURGO data repository. The framework transmits the data on-demand to meet the specific requirements of the user (such as a drone or a biophysical model). The data is published as atomic messages, allowing the user to structure it as desired. The framework integrates with the AVENA framework and uses the NATS message distribution system to communicate among data actors. Integrating public repository data with other data in the IoT system enables data fusion, strengthens data synergy with more contextual detail, and reinforces site-specific agricultural models for real-time decision-making.
Find the publication here: https://elibrary.asabe.org/abstract.asp?JID=5&AID=55000&CID=ana2024&T=1
