A data-driven approach to evaluate emission impacts of manual and automatic transmission vehicles

Understanding and identifying patterns in driver behaviour can yield insights on how to enhance road safety and decrease vehicular emissions. This work focuses on providing a data-driven analysis of the driving behaviour in an urban area, in which each one of the 30 volunteers drives both a manual and automatic transmission vehicle in a simulation experiment. Conducting experiments using a driving simulator can be more cost-effective compared to real-world testing. Different variables are collected (e.g., speed, acceleration, pedal position) and computed or estimated (e.g., vehicle-specific power), and the analysis seeks to explore the patterns in the operational and environmental performance. First results show automatic transmission driving can effectively reduce pollutant emissions, while on the other hand, it can also entail higher and sharper accelerations, more evident at lower speed operating regime.