Systems and methods for high-risk event detection

A system comprising:

• a vehicle monitoring system comprising a sensor configuration including at least one of a vehicle-mounted accelerometer, a mobile device accelerometer, or GPS positioning sensors that receives, in real-time, a telematics data associated with a vehicle collected from one or more data sources, wherein the telematics data includes at least one of a predefined acceleration value over a predefined amount of time or a velocity direction angle relative to a road direction;
• a machine-learning model executed using at least one processor, the machine-learning model trained to identify driving behavior as an occurrence of a high-risk event that exceeds a predetermined threshold value, and continuously generates, in real-time or near real-time, one or more predictions based on the telematics data, wherein the generating comprises correlating the telematics data to a movement indicating a high-risk event using a high-risk prediction algorithm of the machine learning model, wherein the movement is defined by the at least one of the predefined acceleration vector over the predefined amount of time or time series or the velocity vector over the predefined amount of time or time series and determining a numerical score indicating a likelihood of the occurrence of the high-risk event based on the correlation, wherein when the numerical score exceeds the predetermined threshold value the high-risk event is indicated;
• and
• a controller generates a time-based report based on the one or more predictions.

One or more tangible non-transitory computer-readable storage media storing computer-executable instructions for performing a computer process on a computing system, the computer process comprising:

• generating a set of training data, wherein the set of training data includes historical occurrences of events having a designated risk level of high-risk;
• training a machine-learning model to identify high-risk driving behavior using the set of training data;
• receiving, in real-time, a telematics data associated with a vehicle collected from one or more data sources;
• continuously generating in real-time or near real-time, using the machine-learning model, one or more predictions based on the telematics data;
• and
• generating a time-based report based on the one or more predictions, the time-based report identifying instances of the one or more predictions that reach a threshold value,
• wherein the telematics data is received from a sensor configuration including at least one of a vehicle-mounted accelerometer, a mobile device accelerometer, or GPS positioning sensors, wherein the telematics data includes at least one of a predefined acceleration value over a predefined amount of time or a velocity direction angle relative to a road direction, and
• wherein the machine-learning model includes a high-risk prediction algorithm that correlates the telematics data to a movement indicating a high-risk event and determines a numerical score indicating a likelihood of an occurrence of a high-risk event based on the correlation, wherein the movement is defined by the at least one of the predefined acceleration vector over the predefined amount of time or time series or the velocity vector over the predefined amount of time or time series, and wherein when the numerical score exceeds a predetermined threshold value the high-risk event is indicated.