Towards Safer Roads: VSLCloud— A LoRa-Enabled Cloud Platform for Real-Time Alcohol Detection and Vehicle Safety

VSLCloud offers an online, LoRa-compatible cloud-based intoxication detector, an original vehicular safety technology that can prevent the act of driving while intoxicated. Complementing the traditional use of in-car sensors, the VSLCloud system embeds cloud connectivity with IoT-enabled sensors to continuously evaluate intoxication levels and send them to central authorities. It has the capability, in extreme conditions, of calling for alerts or disabling the car by an incapacitated driver. This paper provides an original algorithm, VSLCloud, and compares the same with recognized machine learning-based techniques, i.e., Random Forest (RF), Support Vector Machine (SVM), and Deep Neural Networks (DNN). Test case-based experimental evaluation through ten diverse test scenarios confirms that the VSLCloud surpasses all these techniques by providing the maximum level of accuracy, the fastest response time, and the maximum level of reliability. For example, the VSLCloud achieved a maximum level of accuracy of up to 96.2%, the lowest delay of 120–150 ms, and reliability of over 99% in complex scenarios. These results prove the feasibility of the VSL Cloud-based system of processing real-time information fast enough for timely determination and providing reliable performance through widely fluctuating conditions. The newly designed platform points towards enormous advances towards the development of smart, responsive, and scalable automobile-based safety solutions with great promise toward reducing the incidence of accidents involving intoxication.