• Imen Werda, Tunisia
• Amina Kessentini, Tunisia
• Amna Maraoui, Tunisia

Real-time embedded AI systems combine embedded systems with artificial intelligence to make intelligent decisions within strict time constraints. They are applied across a wide range of industries and domains thanks to their versatility, efficiency, and ability to reliably perform specific tasks. Positioned at the intersection of AI and embedded systems, this rapidly growing field brings intelligent capabilities to a broad spectrum of edge devices and applications. This section explores effective approaches for deploying AI techniques on embedded platforms and highlights recent advances in TinyML hardware–software co-design. It also provides an opportunity to share insights, discuss key challenges, and explore practical solutions for building intelligent systems under strict resource constraints.

Real-time embedded AI systems are widely used in robotics, medical devices, drones, and IoT systems. They are just a few examples of the diverse range of applications where embedded systems play a crucial role. In addition, Designing AI embedded systems involves addressing challenges such as hardware-software co-design, power efficiency, real-time performance, security, robustness to environmental conditions, and scalability across different deployment scenarios.

• Computer Vision: interpretation and analysis of visual data (images & videos)— Object detection & tracking, Facial recognition, Industrial quality control….
• Medical Device: diagnose, monitor or treat medical conditions— AI-powered diagnostic tools: segmentation, classification.
• Smart Home: connected devices to automate and control household systems—IoT, Blockchain.
• Real time IA: Low latency applications, High computational efficiency, Edge deployment.
• Co-design: Hardware software applications— AI model to fit embedded hardware constraints, Balance latency and power consumption.

• Maher Al-Greer, United Kingdom
• Sanjeevikumar Padmanaban, Norway
• Chitra A, India

Batteries are becoming the backbone of electrified mobility, renewable integration, and resilient microgrids. Still, the safe and efficient operation depends on accurate sensing, estimation, and control under uncertainty, aging, and strong electro-thermal nonlinearity. This special session focuses on how artificial intelligence (AI) and modern signal processing can elevate battery management from cell development to pack-level supervision and closed-loop control. We welcome contributions that blend physics-based modeling with data-driven learning, including feature extraction from voltage/current/temperature signals; incremental capacity and differential voltage analysis; impedance pattern extraction; and robust denoising and filtering for embedded deployment.

Key themes include AI-assisted battery development (design-of-experiments, surrogate modeling, and manufacturing quality analytics), advanced state estimation (SOC/SOH/SOP/RUL) with uncertainty quantification, anomaly detection and fault diagnosis (thermal runaway precursors, sensor/actuator faults, cell imbalance), and health-aware optimization and control (fast charging, balancing, thermal management, and predictive energy management). The session also emphasizes trustworthy and deployable AI: interpretability, domain adaptation across chemistries and operating conditions, data governance, and real-time constraints on microcontrollers and automotive-grade BMS hardware. By bringing together researchers from signal processing, control, power electronics, and electrochemistry, the session aims to highlight methods that measurably improve safety, lifetime, and energy efficiency, and to accelerate translation from laboratory benchmarks to fielded battery systems.

• AI/ML for batteries SOC/SOH/SOP/RUL estimation; uncertainty-aware learning.
• Signal processing for battery diagnostics: denoising, ICA/DVA, EIS feature extraction.
• Physics-informed and hybrid electro-thermal models; digital twins for BMS.
• Health-aware control: fast charging, balancing, thermal management, MPC/RL.
• Fault detection & safety: anomaly detection, thermal runaway early warning, cybersecurity.

• Manef BOUROGAOUI, Tunisia
• Zina BOUSSAADA, France

Energy systems are currently undergoing a profound transformation driven by the transition toward renewable, decentralized, and low-carbon energy sources. The continuous increase in the share of solar and wind generation, combined with the development of energy storage technologies and electric mobility, requires traditional power network architectures to evolve toward more intelligent, flexible, and interconnected structures.

This special session addresses the pivotal challenge of integrating renewable energy at high penetration levels into power systems. As electricity networks incorporate growing shares of solar and wind generation, they must evolve from predictable and dispatchable infrastructures into dynamic and variable platforms demanding enhanced intelligence and adaptability across the entire system.

We invite contributions that explore innovative approaches to designing, controlling, and optimizing power systems for a sustainable, resilient, and decentralized future. The session focuses on architecture and methodologies that enable grids to reliably manage renewable generation, ensure stability under high penetration, autonomously balance variable resources, and maintain secure and efficient operation amid growing complexity.

This session seeks to foster a forward-looking mindset that advances solutions for more flexible and grid-ready renewable integration. Contributions on advanced forecasting, variability management, real-time balancing, and grid-strengthening strategies are particularly welcome. The ultimate goal is to build power networks capable of efficiently absorbing high renewable shares, rapidly recovering from disturbances, and fully supporting a clean-energy future.

• Renewable energy generation forecasting.
• Hybrid energy storage management and optimization.
• Intelligent electrical energy consumption management in microgrids using artificial intelligence techniques.
• Electric vehicle integration and bidirectional energy flow management (vehicle-to-grid – V2G).
• Optimization and control methods for hybrid energy systems and microgrids.
• Sizing, planning, and reinforcement of networks with high renewable energy penetration

Renewable Energy Integration; Microgrid Stability and Control; Energy Storage Systems and Hybrid Optimization; Electric Vehicle Integration (V2G); Power System Flexibility and Decentralization; Energy management, Energy production forecasting, Grid Intelligence; Artificial intelligence, Energy Transition

• Inés Abdellaziz, Tunisia
• Férid Chaffar Akkari, Tunisia
• Naoufal Khemiri, Tunisia
• Nabila Elbitri, Tunisia

This special session explores the intersection of artificial intelligence, material science, and advanced sensing to accelerate the transition toward sustainable energy systems. As the demand for high-performance and resilient energy infrastructures grows, the integration of "intelligence" at the physical level becomes crucial. The session focuses on how AI and machine learning algorithms can revolutionize the design, discovery, and characterization of advanced materials with tailored properties for energy storage and conversion. It also highlights the role of smart sensing systems in providing real-time data for health monitoring, predictive maintenance, and optimized performance of energy assets. By bridging the gap between hardware (materials/sensors) and software (AI/Analytics), this session aims to showcase innovative solutions for smarter, more efficient, and self-adaptive energy technologies.

• AI & Machine Learning for Materials Science.
• Data-Driven Modeling & Predictive Analytics.
• Advanced Materials, Nanomaterials & Thin Films.
• Photovoltaic Materials & Energy Harvesting.
• Smart Sensors & Intelligent Sensing Systems.
• Gas Sensors & Sensor Data Analytics
• Energy Monitoring & Digital Twins
• Intelligent Data Fusion for Energy Assets