Fuzzy inference system design for promoting an eco-friendly driving style in IoV domain

Abstract With the rise of both embedded technologies and the Internet of Things (IoT) integration, it is increasingly possible to realize complex systems in which, for example, vehicles are connected to the IoT domain capable of interacting directly with applications on mobile or embedded devices. In this work, a Machine to Machine (M2M) based platform able to recognize the driving style will be presented. An embedded device (On-Board Diagnostic (OBD)) has been considered to implement the Internet of Vehicles (IoV) paradigm exporting recognized data to other systems to promote an eco-friendly driving style by suggesting corrective actions to drivers. The architecture consists of the following elements: an on board device able to obtain information regarding speed, acceleration, jerk and fuel consumption, a cloud able to store all collected data, an M2M protocol such as Message Queuing Telemetry Transport (MQTT) to provide scalability to the architecture, and a Fuzzy Inference System to classify user behavior. All these elements are harmonized to achieve the common target of informing users about an Aggressive or Very-Aggressive driving style with the aim of reducing both speed and acceleration. The smart alerting system will promote a more eco-friendly driving style, contributing a reduction in CO 2 emissions and fuel consumption. Performance evaluation confirms the efficiency of the proposed classifier and the overall benefits for drivers in fuel consumption reduction.