A Tool for Variability Management in Irrigation and Agriculture

Abstract

Agriculture in regions facing severe water scarcity requires highly efficient management strategies, such as deficit irrigation. However, deploying IoT solutions to automate these processes poses a major software engineering challenge due to the immense configuration variability across different commercial farms. To address this, we present Sensolive, an autonomous system for managing olive grove irrigation. Sensolive tackles domain variability by employing a Software Product Line (SPL) architecture via the SPLENT ecosystem, enabling the automated derivation of customized, farm-specific software. Furthermore, it integrates Digital Twins to virtualize physical dendrometer sensors using machine learning, significantly reducing hardware dependency and costs. This tool demonstration showcases Sensolive's capabilities across three phases: estate configuration, digital twin instantiation, and the execution of autonomous deficit irrigation protocols. Sensolive provides a scalable, user-centric architecture that successfully bridges complex agronomic algorithms with customizable deployments.