This paper addresses the challenges of achieving energy efficiency while providing optimal comfort to street users in urban lighting. A system is proposed that uses low-cost embedded devices and an Infrastructure-to-Infrastructure (I2I) paradigm to detect users and adjust illumination levels dynamically. The study introduces a formula called Lighting Comfort Index (LCI) to evaluate user comfort based on the adequacy between provided lighting and user needs. The formula is evaluated through a simulation in Python using different user types (pedestrians and cyclists), varying speeds, and different lighting activation strategies (isolated and communicated). Results highlight the importance of speed and luminaire coordination on perceived comfort. Although preliminary, this simulation-based validation suggests potential use of the formula as a reward function in intelligent lighting control algorithms. Further work includes real-world testing and refinement of parameters.