Centre d'Etudes Superieures Industrielles

Abstract This paper addresses energy‐efficient production scheduling in a flexible flow shop by integrating advanced energy flexibility strategies. A smart manufacturing setting is considered, with energy supplied from multiple sources: grid electricity with time‐of‐use pricing, photovoltaic generation, and an energy storage system. In addition, a demand bidding mechanism allows the facility to submit curtailment capacity bids during periods of power system stress, reducing electricity load in exchange for financial incentives. Two exact optimization approaches are investigated: a mixed integer linear programming model and a hybrid method integrating constraint programming. Computational experiments are conducted on a dedicated benchmark and compared with genetic algorithm, tabu search, and ant colony optimization approaches. Results show that the hybrid method consistently delivers high‐quality solutions, achieving the best relative gap below 1% for all instance sizes, while metaheuristics remain competitive in terms of computation time with small relative gaps.

Heating, ventilation, and air-conditioning (HVAC) systems account for around 16.4% of global final energy consumption and about 14% of global operational CO<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> emissions. Controlling them is a partially observable, sequential decision problem: relying solely on instantaneous sensor readings as inputs overlooks the full sequence of past conditions that shape future dynamics. To tackle this challenge and fully exploit the available temporal context, this study integrates a transformer encoder–decoder architecture into a Double Deep Q-Network (DDQN), forming a "Q-Transformer" capable of processing 24-hour sequences of observations. This approach is benchmarked against a conventional DDQN using a multilayer perceptron (MLP) and a sequence-aware Bi-LSTM within an EnergyPlus model of a 400 m<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> university amphitheater. In a weekly adaptation test, the Q-Transformer demonstrated rapid learning convergence, reducing energy consumption by up to 48% and 66% (occupied and unoccupied periods) compared to Bi-LSTM and MLP, respectively, while maintaining thermal and air quality comfort. After one year of simulated training on weather and occupancy data from a reference site (Luxembourg), the agent was evaluated across 4 different climatic locations, resulting energy reductions of 40% and 74%, and comfort violation reductions of 11% and 29%, compared to Bi-LSTM and MLP, respectively. These results signifies the ability of generalization and control capabilities of transformer-based reinforcement learning for adaptive HVAC management.

This paper introduces an Augmented Perception (AP) framework to enhance robotic perception in resilient manufacturing systems (MS) by integrating Digital Twin (DT) data directly at the sensor level in real time. Inspired by augmented reality, this approach enables robots to perceive both physical and virtual entities within a unified representation. To ensure real-time performance, we address the critical challenges of low-latency data processing and seamless integration. To evaluate the system’s real-time compliance, we propose a latency test and validate our approach through simulations using the Tiago++ robot in Gazebo. Results demonstrate that AP maintains efficient sensor-level integration while meeting real-time processing requirements.

Abstract The integration of Industry 4.0 (I4.0) technologies with Maintenance 4.0 (M4.0) practices holds strong potential for advancing sustainable manufacturing (SM). While these technologies promise improvements in resource efficiency, waste reduction, and alignment with sustainability objectives, research on their synergistic implementation remains limited. This study addresses this gap through a Systematic Literature Review (SLR) of 75 peer-reviewed papers published between 2015 and 2024, conducted in accordance with PRISMA guidelines. The review explores how I4.0–M4.0 synergy contributes to sustainability across four interconnected dimensions: economic, environmental, social, and technological. Findings show that integrating technologies such as the Internet of Things and Artificial Intelligence into maintenance operations can reduce downtime by 20–50% and enhance efficiency and system resilience by 10–25%, particularly in industries like automotive and aerospace. Digital twin technologies extend equipment lifespan by 10–25%, thereby deferring capital expenditures. Furthermore, blockchain and augmented reality improve operational transparency by 30–40%, while big data analytics and cyber-physical systems contribute to energy savings of 12–18% and reduce material waste by 20–25% through real-time quality monitoring. Despite these benefits, several challenges hinder integration, including technical barriers (e.g., legacy systems, cybersecurity risks), organizational resistance (e.g., high costs, cultural inertia), and human-related issues (e.g., skills shortages, workforce restructuring). To address these barriers, the paper proposes a holistic architecture that aligns I4.0–M4.0 integration with sustainability goals, bridging technological innovation with responsible resource management. This framework offers actionable insights for stakeholders, policymakers, and industry leaders aiming to foster resilient, efficient, and socially responsible manufacturing ecosystems.

Detecting impact where an individual makes contact with the ground within a fall event is crucial in fall detection systems, particularly for elderly care where prompt intervention can prevent serious injuries. The UP-Fall dataset, a key resource in fall detection research, has proven valuable but suffers from limitations in data accuracy and comprehensiveness. These limitations cause confusion in distinguishing between non-impact events, such as sliding, and real falls with impact, where the person actually hits the ground. This confusion compromises the effectiveness of current fall detection systems. This study presents enhancements to the UP-Fall dataset aiming at improving it for impact fall detection by incorporating 3D skeleton data. Our preprocessing techniques ensure high data accuracy and comprehensiveness, enabling a more reliable impact fall detection. Extensive experiments were conducted using various machine learning and deep learning algorithms to benchmark the improved 3D skeletons dataset. The results demonstrate substantial improvements in the performance of fall detection models trained on the enhanced dataset. This contribution aims to enhance the safety and well-being of the elderly population at risk.

Online continuous action recognition has emerged as a critical research area due to its practical implications in real-world applications, such as human-computer interaction, healthcare, and robotics. Among various modalities, skeleton-based approaches have gained significant popularity, demonstrating their effectiveness in capturing 3D temporal data while ensuring robustness to environmental variations. However, most existing works focus on segment-based recognition, making them unsuitable for real-time, continuous recognition scenarios. In this paper, we propose a novel online recognition system designed for real-time skeleton sequence streaming. Our approach leverages a hybrid architecture combining Spatial Graph Convolutional Networks (S-GCN) for spatial feature extraction and a Transformer-based Graph Encoder (TGE) for capturing temporal dependencies across frames. Additionally, we introduce a continual learning mechanism to enhance model adaptability to evolving data distributions, ensuring robust recognition in dynamic environments. We evaluate our method on the SHREC'21 benchmark dataset, demonstrating its superior performance in online hand gesture recognition. Our approach not only achieves state-of-the-art accuracy but also significantly reduces false positive rates, making it a compelling solution for real-time applications. The proposed system can be seamlessly integrated into various domains, including human-robot collaboration and assistive technologies, where natural and intuitive interaction is crucial.

International audience