Weather-based road condition estimation in the era of Internet-of-Vehicles (IoV)

Environmental-Based Speed Recommendation for Future Smart Cars

“…The impact of vehicle connectivity is considered crucial in terms of vehicle safety and, for that reason, the National Highway Traffic Safety Administration (NHTSA) of the USA expects that every new car will be equipped with connectivity capabilities after year 2020 [12]. Thus, smart cars are becoming the most sophisticated entity in the Internet of Things (IoT) era, hence they prevail as a replacement to modern cars introducing the era of Internet of Vehicles [13][14][15].…”

“…For instance, camera systems have difficulty in detecting bright objects when the camera sensor is blinded by the sun, whereas a radar system is able to detect other vehicles or obstacles in the same situation. As a result, data are susceptible to imprecision and inconsistency [14,16] and making single information source approaches inefficient and error prone, as they can lead to imprecise results. Therefore, it becomes prominent to abandon traditional single source approaches and move on multi source strategies.…”

“…As a use case, the presented methodology combines information from multiple heterogeneous sources and alerts the driver about the recommended speed that the car should follow, so that it can avoid dangerous driving. Compared to our original work presented in [14], the differentiators of the presented methodology are the following:…”

“…We further explain the details of our original implementation in [14], providing more details about the data analysis (Section 3.2). We also evaluate and compare the presented classification model against existing techniques and we present the results in Section 4.4.…”

“…We developed the overall service on top of Hydra [20], a distributed multi-agent computing framework that targets smart cars and it provides self-management functions such as in-car workload balancing and failure recovery. Section 4.1 further explains the implementation presented in [14] and enhances this work with more technical details regarding the development of the proposed methodology on real computing boards by utilizing the computation infrastructure in [20].…”