Smart Cyber-Physical Systems Dataset for Real-Time Optimization, Resilience, and Sustainability in Industry 5.0

This dataset provides a large-scale, scientifically grounded representation of smart cyber-physical systems (CPS) designed for real-time optimization in Industry 5.0 contexts. It comprises 500 CPS instances across mobility, energy, industrial, and urban infrastructure domains, integrating multi-dimensional indicators related to sensing density, latency, control stability, optimization efficiency, and system resilience. The dataset combines artificial intelligence and advanced control perspectives by including variables associated with reinforcement learning performance, model predictive control constraint violations, digital twin fidelity, edge and cloud resource utilization, and learning convergence behavior. In addition, it captures critical dimensions of cybersecurity, trust, and governance, such as cybersecurity events, blockchain-based verification latency, data integrity, ethical compliance, and human-in-the-loop participation. Sustainability and human-centric performance are explicitly addressed through energy consumption, carbon emissions, carbon awareness, and composite sustainability indicators. The dataset is suitable for empirical analysis, simulation-based validation, machine learning model training, and comparative studies on real-time optimization, resilient CPS architectures, and human-centered AI systems, supporting reproducible research and decision-making in smart infrastructures and Industry 5.0 applications.