Titre :	IOT based smart irrigation management system
Type de document :	document multimédia
Auteurs :	Nour El Houda Djekidel, Auteur ; Fatima Zahra Lahreche, Auteur ; Lakhdar Kamel Oulad Djedid, Directeur de thèse
Editeur :	Laghouat : Université Amar Telidji - Département d'informatique
Année de publication :	2026
Importance :	65 p.
Accompagnement :	1 disque optique numérique (CD-ROM)
Note générale :	Option : Networks, distributed systems and applications
Langues :	Anglais (eng)
Mots-clés :	Smart irrigation  AIoT  XGBoost  Hybrid Edge–Cloud  Concept drift  Precision agriculture  Water management
Résumé :	The agricultural sector faces increasing challenges due to water scarcity and the inefficiency of traditional irrigation systems, which rely on fixed thresholds rather than accurate environmental data, leading to water wastage and suboptimal crop yields. To address these issues, this study proposes an adaptive hybrid Edge--Cloud AIoT smart irrigation system integrating ESP32-based IoT sensors with ML real-time irrigation decisions. A comparative evaluation of six classification models was conducted, with XGBoost identified as the optimal model achieving an F1 score of 0.896 (∼89.7%) prediction accuracy across three irrigation demand classes. The system features a dynamic routing mechanism with an adaptive confidence threshold (automatically updated from a sliding window of the last 20 predictions using mean and standard deviation statistics) that assigns high-confidence tasks to a lightweight edge rule engine while escalating complex cases to the cloud-based model. The system processes approximately 26--30% of requests locally at the edge layer, reducing cloud dependency and inference latency. An adaptive concept drift detection module further triggers autonomous background retraining to maintain long-term system robustness.
note de thèses :	Mémoire de master en informatique

IOT based smart irrigation management system [document multimédia] / Nour El Houda Djekidel, Auteur ; Fatima Zahra Lahreche, Auteur ; Lakhdar Kamel Oulad Djedid, Directeur de thèse . - Laghouat : Université Amar Telidji - Département d'informatique, 2026 . - 65 p. + 1 disque optique numérique (CD-ROM).
Option : Networks, distributed systems and applications
Langues : Anglais (eng)

Mots-clés :	Smart irrigation  AIoT  XGBoost  Hybrid Edge–Cloud  Concept drift  Precision agriculture  Water management
Résumé :	The agricultural sector faces increasing challenges due to water scarcity and the inefficiency of traditional irrigation systems, which rely on fixed thresholds rather than accurate environmental data, leading to water wastage and suboptimal crop yields. To address these issues, this study proposes an adaptive hybrid Edge--Cloud AIoT smart irrigation system integrating ESP32-based IoT sensors with ML real-time irrigation decisions. A comparative evaluation of six classification models was conducted, with XGBoost identified as the optimal model achieving an F1 score of 0.896 (∼89.7%) prediction accuracy across three irrigation demand classes. The system features a dynamic routing mechanism with an adaptive confidence threshold (automatically updated from a sliding window of the last 20 predictions using mean and standard deviation statistics) that assigns high-confidence tasks to a lightweight edge rule engine while escalating complex cases to the cloud-based model. The system processes approximately 26--30% of requests locally at the edge layer, reducing cloud dependency and inference latency. An adaptive concept drift detection module further triggers autonomous background retraining to maintain long-term system robustness.
note de thèses :	Mémoire de master en informatique