About RoadSafe Assistant

The RoadSafe Assistant system is a comprehensive video monitoring solution tracks driver behaviour and condition consisting of both hardware and software based on advanced technologies. Its primary goal is to significantly reduce accidents caused by driver fatigue, ensuring safer roads and enhancing overall transportation safety.

★ The AI-driven assistant provides signals through visual, auditory, and voice cues to ensure real-time intervention and alerts the dispatch analysis centre, preventing accidents on the roads and enhancing safety.

★ Additionally, it suggests breaks based on real-time fatigue detection and engages the driver with music or specially elaborated questions to maintain alertness and focus.

★ The system is designed to detect indicators of potentially dangerous events.

Why is it important ?

The factors contributing to the increase in the number of accidents:

★ speeding,

★ alcohol or drugs,

★ fatigue,

★ distracted driving,

★ not wearing a seatbelt, etc.

However, the main factors are: drowsiness and distracted driving.

They mostly lead to road accidents and endanger drivers, passengers, pedestrians, and fellow motorists.

Technologies the assistant is based on

Computer Vision: capturing and analysing the camera stream to detect driver's states and behaviour;

-Artificial Intelligence: intelligent data analysis, recognition and classification of risky driver behaviour signals;

-Special Machine Learning Algorithms: training a machine based on accumulated data to improve risk detection accuracy;

-Multilayer Neural Networks: deep video and data analysis, identifying complex patterns and correlations;

-Highly Sensitive Sensors: installed in the camera, focusing on dozens of facial points to detect danger signals;

-Natural Language Understanding: recognising and processing driver commands and requests;

-Speech synthesis: generating voice alerts and feedback for a driver;

-Edge Computing: processing data straight to the vehicle's on-board device for immediate response;

-Internet of Things (IoT): data exchange between the vehicle's on-board device, cloud platform and monitoring centre;

-Infrared Camera: visualization of the driver's condition in dark or bad weather conditions;

-Smart Display: displaying alerts and information to attract the driver's attention;

-Central Processing Unit: video stream is sent to the computing unit, transmitting telemetric and geolocation data to the neural network.

What dangers our system can detect

-Unfastened seatbelt;

-Talking on a mobile phone;

-Smoking while driving;

-Drowsiness or fatigue (yawning, closed eyes, frequent blinking);

-Absence of driver behind the wheel;

-Obstruction of the camera or lens deflection;

-Signs of lost concentration (looking away, distractions from a mobile phone);

-Rapid eye movement indicative of microsleeps;

-Slouching or poor posture signaling physical fatigue;

-Excessive lane deviation or swerving;

-Customizable recognition settings for client-specific alert events is possible due to our special learning algorithms.

How it works?

1.Once the vehicle engine is running, the camera is automatically switched on and analyses the driver's gesticulation throughout the whole trip.

2. In case of warning signs, the System automatically alerts the driver with an audible or visual signal.

3.The system suggests breaks based on real-time fatigue detection and engages the driver with specially elaborated questions to maintain alertness and focus.

4. In addition to the alert, the software notifies the dispatch analysis centre of the incident and indicates the data on a potential danger event.

5.Multilayer Neural Networks are used by dispatchers at the analysis center to validate possible accidents.

6.If needed, the dispatch analysis centre sends a message to the representative of the company owning the vehicle fleet, mentioning the emergency event.

7.The client has at his disposal all the required data: statistics or analytics on the fleet both in real time and in any past period.