Défense de Thèse de Doctorat en “Electronics : Embedded Systems and Artificial Intelligence” by Ms. Nisrine LACHGAR

The Euromed University of Fes (UEMF) is pleased to inform the public of

the doctoral thesis defense in “Electronics: Embedded Systems and Artificial Intelligence”

The thesis defense will take place on December 1, 2025 at 9:00 a.m. at UEMF

Location: The Great Hall of the Incubator (LOC001994)

The thesis will be presented by Ms. Nisrine LACHGAR

under the topic:

 “Advanced Precision Irrigation Through AI-based Modelling and Expert Systems for Embedded Environments”

Summary

Climate change, water scarcity, and rising agricultural demands are placing unprecedented pressure on irrigation practices, particularly in water-stressed regions. Precision irrigation has emerged as a promising solution, yet its potential remains limited by fragmented systems lacking real-time intelligence, adaptability, and explainability. This thesis addresses these challenges by proposing an integrated decision-support framework that combines artificial intelligence, semantic reasoning, and embedded system design to support sustainable, data-driven irrigation strategies.

The research builds upon a dual analytical foundation: a systematic review of the literature on Artificial Intelligence models for evapotranspiration estimation and the technological landscape of embedded systems in agriculture. This synthesis revealed both potential and limitations in current approaches, underscoring the need for a unified architecture that merges predictive modeling with actionable, knowledge-based automation. Based on this foundation, a modeling pipeline is developed using climatic data from the Fez region. After rigorous preprocessing, various learning models are examined for reference evapotranspiration (ET₀), with time series methods demonstrating better adequacy due to their ability to capture temporal dependencies in meteorological patterns. A sensitivity-based input reduction strategy is introduced to ensure efficient implementation under hardware and data constraints. To complement the predictive layer, a modular expert system is developed using ontology-based knowledge representation and rule-based inference. By encoding agronomic and environmental knowledge into formal logic, the system delivers context-aware, explainable irrigation recommendations aligned with local decision constraints.

The thesis culminates in the proposal of a scalable embedded system architecture integrating environmental monitoring, on-device reasoning, and actuation. This approach bridges scientific modeling with real-world deployment, laying the foundation for adaptive, explainable irrigation control. Future work will extend the framework through dynamic crop coefficient modeling, cross-regional validation, and real-time implementation in operational agricultural settings.

This thesis will be presented to the jury members:

Partager