Dynamic Street Lighting based on Adaptive Learning

Smart street light solution utilizing Long Range Wide Area Network (LoRaWAN) is one of the highly interesting topics in the field of Internet of Things (IoT) automation engineering. It is seen that many prototypes and modelling were served with IoT based sensors in this existing research idea globally and in fact Narrow Band IoT is used in some cases for data sharing along with secure communication channels. But some factors other than sensory technology associated with it will become more crucial to consider as those are emerging now like street light controlling as per traffic loads, adjustment of illumination power dynamically, and some changes of road condition if possible to distribute the data under a fault-tolerant architecture. In this paper, there is proposed an adaptive algorithm for street light control with decision-making logic based on Reinforcement Learning in some of the road cases as discernment. And with it traffic pattern observation to feed the dataset further using a deep learning technique. Also, simulate a Deep Q-learning algorithm that is necessary to be shown here and the most enthusiastic thing of implementation is to observe the road patterns and vehicular movements and note them clearly.