Research in autonomous vehicles has recently attracted strong interest from both industry and academia. While there have been significant advances in this area for external sensing, path planning, and vehicle control of autonomous vehicles, to the best of our knowledge few studies have addressed the driving efficiency of autonomous cars. On the other hand, an eco-driving advisory system (EDAS), which leverages the signal phase and timing (SPAT) information, can provide autonomous vehicles with eco-driving suggestions and minimize the stops for signals, resulting in a reduction in CO2 emissions and travel time. In this work, we proposed a novel speed advisory system named the enhanced maximized throughput model (EMTM) by addressing the shortcomings of the prior work such as MaxTM and MinADM. The simulation results show that the proposed EMTM model is up to 13.6% better than MaxTM and 26.3% better than MinADM with regard to CO2 emissions.