Bätz, AlexanderAlexanderBätzGimpel, HennerHennerGimpelHeger, SebastianSebastianHegerWöhl, MoritzMoritzWöhl2022-03-142022-03-142020https://publica.fraunhofer.de/handle/publica/40875310.24251/HICSS.2020.1142-s2.0-85102354416One starting point to reduce harmful greenhouse gas emissions is driving behavior. Previous studies have already shown that eco-feedback leads to reduced fuel consumption. However, less has been done to investigate how driving behavior is affected by eco-feedback. Yet, understanding driving behavior is important to target personalized recommendations to-wards reduced fuel consumption. In this paper, we investigate a real-world data set from an IoT-based smart vehicle service. We first extract seven distinct factors that characterize driving behavior from data of 5,676 users. Second, we derive initial hypotheses on how eco-feedback may affect these factors. Third, we test these hypotheses with data of another 495 users receiving eco-feedback. Results suggest that eco-feedback, for instance, reduces hard acceleration maneuvers while interestingly speed is not affected. Our contribution extends the understanding of measuring driving behavior using IoT-based data. Furthermore, we contribute to a better understanding of the effect of eco-feedback on driving behavior. One starting point to reduce harmful greenhouse gas emissions is driving behavior. Previous studies have already shown that eco-feedback leads to reduced fuel consumption. However, less has been done to investigate how driving behavior is affected by eco-feedback. Yet, understanding driving behavior is important to target personalized recommendations towards reduced fuel consumption. In this paper, we investigate a real-world data set from an IoT-based smart vehicle service. We first extract seven distinct factors that characterize driving behavior from data of 5,676 users. Second, we derive initial hypotheses on how eco-feedback may affect these factors. Third, we test these hypotheses with data of another 495 users receiving eco-feedback. Results suggest that eco-feedback, for instance, reduces hard acceleration maneuvers while interestingly speed is not affected. Our contribution extends the understanding of measuring driving behavior using IoT-based data. Furthermore, we contribute to a better understanding of the effect of eco-feedback on driving behavior.eneco-feedbackdriving behaviorreal-world datafactor analysis004005006conference paper