Internet of Vehicles (IoV) based Framework for Vehicle Degradation using Multidimensional Dynamic Time Warping (MDTW).

The anomalies on highways impose wear-tear which may cause degradation in prolonging of the vehicles. Although, past research has played an important role in detecting potholes and speed bumps with high accuracy. This study proposes an Internet of Vehicles (IoV) based framework for detecting road anomalies using Multidimensional Dynamic Time Warping (MDTW) for improving the F1 score and sensitivity. Furthermore, the vehicle degradation index caused by road anomalies is established. The system employs a Three-layer architecture comprising IoV, Fog, and cloud computing, utilizing their inherent capabilities as needed for the specific task at hand. The data is gathered using the IoV system, which includes an Arduino MEGA 2560 equipped with an ESP8266, M8N GPS, and MPU-6050. The fog layer computes the vehicle's jerk impact using onboard capabilities and a complementary filter for bandwidth optimization. The cloud layer, with powerful computing capabilities, evaluates the proposed MDTW method against existing techniques for detecting potholes and speed bumps, serving as a potential backbone for the model. The proposed approach achieved an F1 score of 0.8934 and 0.8852 for the studied datasets. A correlation was found between prolonging the lifespan of a vehicle and the frequency of jerks experienced during two trips, as indicated by the Vehicle Degradation Index with an average score of 0.7212. To summarize, the proposed framework can be used not only in the automobile industry, where production can be customized based on demographic factors but also in low-cost automatic monitoring and repair of roads.