Corrigendum to “Fuel economy testing of autonomous vehicles” [Transp. Res. Part C: Emerging Technol. 65 (2016) 31-48]

This paper presents a reanalysis of two of the models and one of the vehicles used in a previously published paper “Fuel Economy Testing of Autonomous Vehicles” (Mersky and Samaras, 2016). In that paper, a method for testing the fuel economy of automated vehicle features was proposed and tested, via simulation. In that paper, for the rule-sets and vehicles tested, it was estimated that fuel economy could change from −3 to +10%; this has been amended to −3 to +5%. In the prior paper, it was concluded that the VelocityACC control method occasionally improved fuel economy, that HeadwayACC generally performed worse than VelocityACC and that PlannedACC performed best. HeadwayACC and VelocityACC were also found to be insensitive to target headway. In this paper implementation errors were corrected in VelocityACC and HeadwayACC. Additionally, of the four vehicles tested, the 2011 Toyota Camry was rerun with four cylinders, rather than six. Under the assumptions used, VelocityACC has been found to never increase fuel economy, when following base EPA cycles. HeadwayACC has been found to always increase fuel economy and perform better than PlannedACC, when following base EPA cycles. Both HeadwayACC and VelocityACC were found to be sensitive to desired headway, improving fuel economy as desired headway increased. These amended results serve to further reinforce the importance of testing automated vehicle feature fuel economy as was concluded in the previous paper.