Safe driving has become a critical necessity in the current scenario, given the sharp rise in vehicle number coupled with slow development of road infrastructure. Automated vehicles equipped with cloud connectivity appear to have only a limited impact on easing traffic congestion. Issues such as accidents, traffic congestion, and driving safety are in need of significant improvements. In this work, we propose to collect the necessary data from all the vehicles, process, derive appropriate control signals and send to all the vehicles. Meanwhile, each vehicle is designed to communicate its status to all other vehicles within its range, thereby, improving the driving conditions. The On-Board Unit is developed with Seeed Studio XIAO ESP32C3 along with sensors to extract eight features from every vehicle. The data collected by the module is processed using machine learning [ML] algorithms. Wi-Fi and Bluetooth capabilities of microcontrollers are exploited for V2V and V2X communication. It was observed that among logistic regression, Naïve Bayes, K-NN, SVM, and DT classification algorithms, DT was more suitable for this application with classification accuracy at 97%. The accuracy achieved by the low-cost, lightweight module is highly promising compared to previous studies. An additional advantage is that the module can also be used by pedestrians to enhance their safety around vehicles.