Energy Efficient Scheduler Based on GRU Model for Smart Street Light Using NBIoT

Smart street lighting systems, heralded for their potential to revolutionize urban infrastructure, are increasingly embraced by cities aiming to enhance energy efficiency and operational monitoring. However, these systems are not without their challenges, particularly concerning energy management. Therefore, an energy-efficient scheduler using an optimized GRU (Gated Recurrent Units) model is proposed in this paper to reduce energy consumption through adaptive lighting based on environmental conditions and pedestrian traffic. Narrowband Internet-of-Things (NBIoT) protocol, which is a very low-power protocol based on cellular technology that leverages existing LTE infrastructure to enable a massive number of connected devices, is also integrated into our approach for remotely controlling the brightness of the street light system. By using an AI GRU model to generate tailored lighting scenarios for each street, the system achieves significant energy savings up to 80%, while only reducing brightness by 45% during the night-time. This level is likely sufficient for general street lighting needs contributing to energy conservation without drastically compromising visibility.