Car accidents caused by drowsiness present a concerning problem within the Brazilian traffic context, with substantial implications for road safety and public well-being.
According to available data, car accidents related to drowsiness have been alarmingly prevalent in Brazil. The National Traffic Department (Departamento Nacional de Trânsito - DENATRAN) reported that driver fatigue accounted for a significant proportion of road accidents across the country. In fact, a study conducted by the Brazilian Association of Road Traffic Medicine (Associação Brasileira de Medicina de Tráfego - ABRAMET) revealed that drowsiness contributed to approximately 42% of all recorded car accidents in Brazil. Another 18% are related to excessive fatigue.
Aware of all of these problems, we are promoting AutoFleet. A device that utilizes cutting-edge technology to detect signs of drowsiness and provide timely alerts to keep truck drivers awake and focused behind the steering-wheel, as well as providing reports about each trip containing speed, acceleration and offenses data to its manager. It's important to note that in the AutoFleet context, offenses are caused when the vehicle speed is excessive, when the driver shows signs of fatigue or when the driver keeps his hands out of the wheel for a long time.
To be able to fetch and process all this data, our system uses a combination of cameras, gps and accelerometer sensors combined with intelligent algorithms embedded into a raspberry PI to monitor key parameters such as eye movements, head and hands position, vehicle speed and acceleration and driver behavior overall. When fatigue or hands off the wheel are detected, AutoFleet instantly trigger a buzzer calling the driver's attention ensuring he will remain attentive and alert throughout his journey.
Not only this, but also all the information of each trip is sent regularly as packets to a platform hosted in a cloud server which can be checked later on by the fleet manager. It's important to recall that these packets are only sent if the device has internet connection. However, if the connection is lost, the device manages to store these packets locally in a micro-sd card embedded in the microprocessor until the connection is reestablished.
The design architecture of the system is better explained in this chart, which shows the relation of all the sensors mentioned above within the raspberry, and the raspberry relation with the cloud server. It's important to notice that all the data is stored in a cloud database inside AWS, where the server is hosted as well.
About the platform interaction, both the manager and the driver hold a major role. The driver needs to log in, start and end a new trip in order to keep the device able to communicate with the server. On the other hand, the fleet manager has an overall overview of each driver, being able to access each trip to check if any offenses occurred.
In conclusion, the introduction of AutoFleet holds great promise in addressing the problem of drowsy driving, and bad behavior aiming to provide a better and safer traffic for everyone.