This paper presents a comprehensive analysis of the impact of automated driving on energy consumption in real-world driving conditions based on a large-scale dataset of driving data from a fleet of diverse vehicles and drivers. In highlighting the current energy efficiency differences between automated drivers and human drivers, this study establishes a baseline from which the behavior of automated driving systems can be improved to achieve widespread energy savings in the transportation sector.

Together with members of the Vehicle Mobility Systems team at Argonne National Laboratory, including Ayman Moawad (lead), Jihun Han, Dominik Karbowski, Yaozhong Zhang, and Aymeric Rousseau, Matt Zebiak from 2050 Partners explores the root causes behind observed efficiency differences between automated drivers and human drivers. Based on these results, the team offers tangible recommendations for vehicle manufacturers and automated driving software companies to achieve better energy efficiency in their automated driving systems, from basic cruise control to fully autonomous vehicles.

The full text is available for free at https://www.nature.com/articles/s41467-024-54066-8.