An Amalgamation of Machine Learning and Embedded Systems for Smart Street Lightning Systems

This research paper introduces an innovative Smart Street Light System that optimizes street lighting operations by harnessing ambient light levels to govern the activation and deactivation of streetlights. Lux value thresholds are employed to distinguish between various lighting scenarios, ranging from bright sunlight to subdued sunlight and complete darkness. Moreover, the paper explores the potential integration of an ultrasonic sensor (US 100) for detecting approaching vehicles and pedestrians, by using advanced machine learning algorithms, thereby enhancing system capabilities. The novel use of Machine Learning algorithms, extensive training of the model, on dataset built using real world conditions, enhances the output accuracy of the model, which gives the model the capabilities to take decisions on its own and thus create a self-sustainable ecosystem, harnessing advanced technologies, without any human involvement. The model was created and implemented with an accuracy of 97.35 percent.