Ubiquitous sensorization for multimodal assessment of driving patterns

Abstract

Sustainability issues and sustainable behaviours are becoming concerns of increasing signi cance in our society. In the case of transportation systems, it would be important to know the impact of a given driving behaviour over sustainability factors. This paper describes a system that integrates ubiquitous mobile sensors available on devices such as smartphones, intelligent wristbands and smartwatches, in order to determine and classify driving patterns and to assess driving e ficiency and driver's moods. It first identi fies the main attributes for contextual information, with relevance to driving analysis. Next, it describes how to obtain that information from ubiquitous mobile sensors, usually carried by drivers. Finally, it addresses the multimodal assessment process which produces the analysis of driving patterns and the classi cation of driving moods, promoting the identifi cation of either regular or aggressive driving patterns, and the classi fication of mood types between aggressive and relaxed. Such an approach enables ubiquitous sensing of personal driving patterns across diff erent vehicles, which can be used in sustainability frameworks, driving alerts and recommendation systems.