Embedding Environmental Intelligence into Digital Twins for Resource-Aware Process Control in Computer Networks

Digital infrastructure contributes significantly to global electricity consumption, with data centres, high-speed communication networks, and edge devices operating continuously to meet growing computing demands. The infrastructure is evaluated based on performance characteristics such as throughput, latency, and fault tolerance. This study proposes a novel framework integrating environmental intelligence into digital twins to enable resource-aware process control in digital infrastructure. Factors including power usage, temperature, and e-waste generation are incorporated as key components. The digital twin model uses the energy profiles of routers, switches, and computing nodes across time and usage conditions, generating real-time data to predict variations and impacts. A multi-objective optimisation engine was developed using a weighted-sum approach to balance sustainability and performance objectives, with constraints on SLA adherence and hardware availability. The objective function optimises performance and energy consumption while maintaining network performance. We designed a proof-of-concept framework that acts like a cloud-edge network. The results showed that applying the modelling resulted in a 12.6% reduction in energy consumption and a 9.8% increase in performance under typical load scenarios. The system dynamically rerouted non-critical traffic during peak grid emissions, activated low-power modes during idle periods, and recommended infrastructure upgrades based on thermal hotspot forecasts and energy impact assessments. The proposed framework demonstrates how digital twins can align operational efficiency with sustainability by embedding intelligence into real-time control mechanisms. This approach supports the broader vision of intelligent and responsible infrastructure management for next-generation computing systems.