An urban smart lighting architecture allowing municipalities to manage and control public street lighting lamps is here considered. The system is capable to autonomously modulate the street lamps’ brightness in specific areas or segments of the urban roads of interest to reduce the energy consumption at the minimum allowed by regulations. To this end, a hybrid traffic model is proposed in this article to estimate in each segment of the roads the number of vehicles on which basis the lamps’ brightness minimization is accomplished. The traffic model represents a digital twin of the smart architecture that acts by orchestrating local dimmers, traffic lights, video cameras, and acoustic sensors. A bank of robust switching observers is used to estimate and classify the traffic in all segments. One agent at a time is selected from the bank and put in order. The agent selection is based on a distributed reputation mechanism recently developed for a class of networked systems. As one of its main merits, this methodology is capable to reveal those sensors that have the highest informative contents (Quality of Service) on the basis of their reputation (trust) shared among the agents. Finally, the proposed approach has been evaluated by considering a highly structured road network and, as one of its main results, the simulations show energy savings up to 70% with respect to the traditional street lamps management.