System and method for predicted vehicle incident warning and evasion

1. A data collection and processing device in a vehicle configured to provide predicted vehicular event warning and/or evasion, the device comprising:

at least one of a plurality of sensors and a signal interface to the plurality of sensors;

a memory configured to store a plurality of instructions and a pre-trained pattern recognition algorithm; and

a processor, coupled to the plurality of sensors via the signal interface and to the memory, wherein the processor is configured to execute the plurality of instructions to:

receive real-time data from the plurality of sensors relating to at least one of a current vehicle condition and a current driver condition,

predict, by processing at least a portion of the real-time data through the pre-trained pattern recognition algorithm, a likelihood that a predefined vehicle event from among a plurality of predefined vehicle events will occur,

wherein the pre-trained pattern recognition algorithm comprises a pattern recognition module that is trained to predict the likelihood that the predefined vehicle event will occur using: (a) previously-collected event data, which corresponds to data that characterizes prior occurrences of the predefined vehicle event, and (b) previously-collected interval data, which corresponds to data collected at predetermined time intervals preceding the predefined vehicle event,

wherein the event data is collected in response to detecting the predefined vehicle event, whereas the interval data is collected at the predetermined time intervals independent of detecting the prior occurrences of the predefined vehicle event, and

wherein the pre-trained pattern recognition algorithm predicts the likelihood that the predefined event from among the plurality of vehicle events will occur via pattern matching the real-time data to the previously-collected interval data,

output a first warning type if the likelihood is predicted to be above a first threshold but not above a second threshold, and

output a second warning type and/or conducting a vehicle evasive maneuver if the likelihood is predicted to be above the second threshold.

2. The device of claim 1 , wherein the previously-collected event data is collected over a predetermined window of time that corresponds to a time of the prior occurrence of the at least one predefined vehicle event.

3. The device of claim 1 , wherein at least one of the previously-collected event data and previously-collected interval data is collected from a plurality of vehicles driven by a plurality of drivers.

4. The device of claim 1 , wherein the pre-trained pattern recognition algorithm is configured to perform a pattern recognition operation on the received real-time data using a set of target values corresponding to the at least one predefined vehicle event.

5. The device of claim 1 , wherein the second warning type comprises at least one of (i) illuminating a symbol on a display panel in a more pronounced manner than in the case of the first warning type, (ii) outputting an audible sound from a speaker of the vehicle in a more pronounced manner than in the case of the first warning type, and (iii) providing a haptic warning.

6. The device of claim 1 , wherein conducting the vehicle evasive maneuver comprising at least one of (i) controlling a brake pedal or braking system of the vehicle to avoid or mitigate the predicted at least one predefined vehicle event, (ii) controlling a steering wheel and/or steering system of the vehicle to avoid or mitigate the predicted at least one predefined vehicle event, and (iii) controlling a vehicle throttle, accelerator pedal or steering angle to avoid or mitigate the predicted at least one predefined vehicle event.

7. The device of claim 1 , wherein the pre-trained pattern recognition algorithm is ported to the memory of the device from a backend server after undergoing a database reduction operation.

8. The device of claim 1 , wherein the real-time data further comprises real-time data relating to at least one of environmental conditions and driving conditions.

9. A method for providing predicted vehicular event warning and evasion in a vehicle, the method comprising:

receiving, by a data collection and processing device in a vehicle from a plurality of sensors, real-time data relating to at least one of a current vehicle condition and a current driver condition;

predicting, by processing at least a portion of the real-time data through a pre-trained pattern recognition algorithm stored in a memory of the vehicle, a likelihood that a predefined vehicle event from among a plurality of predefined vehicle events will occur,

wherein the pre-trained pattern recognition algorithm comprises a pattern recognition module that is trained to predict the likelihood that the predefined vehicle event will occur using: (a) previously-collected event data, which corresponds to data that characterizes prior occurrences of the predefined vehicle event, and (b) previously-collected interval data, which corresponds to data collected at predetermined time intervals preceding and the predefined vehicle event,

wherein the event data is collected in response to detecting the predefined vehicle event, whereas the interval data is collected at the predetermined time intervals independent of detecting the prior occurrences of the predefined vehicle event, and

wherein said predicting is via pattern matching the real-time data to the previously-collected interval data; and

outputting a first warning type if the likelihood is predicted to be above a first threshold but not above a second threshold; and

outputting a second warning type and/or conducting a vehicle evasive maneuver if the likelihood is predicted to be above the second threshold.

10. The method of claim 9 , wherein the previously-collected event data is collected over a predetermined window of time that corresponds to a time of the prior occurrence of the at least one predefined vehicle event.

11. The method of claim 9 , wherein at least one of the previously-collected event data and previously-collected interval data is collected from a plurality of vehicles driven by a plurality of drivers.

12. The method of claim 9 , further comprising performing, by the pre-trained pattern recognition algorithm, a pattern recognition operation on the received real-time data using a set of target values corresponding to the at least one predefined vehicle event.

13. The method of claim 9 , wherein the second warning type comprises at least one of (i) illuminating a symbol on a display panel in a more pronounced manner than in the case of the first warning type, (ii) outputting an audible sound from a speaker of the vehicle in a more pronounced manner than in the case of the first warning type, and (iii) providing a haptic warning.

14. The method of claim 9 , wherein conducting the vehicle evasive maneuver comprising at least one of (i) controlling a brake pedal or braking system of the vehicle to avoid or mitigate the predicted at least one predefined vehicle event, (ii) controlling a steering wheel and/or steering system of the vehicle to avoid or mitigate the predicted at least one predefined vehicle event, and (iii) controlling a vehicle throttle or accelerator pedal to avoid or mitigate the predicted at least one predefined vehicle event.

15. The method of claim 9 , further comprising porting the pre-trained pattern recognition algorithm to a memory of the data collection and processing device from a backend server after undergoing a database reduction operation.

16. The method of claim 9 , wherein the real-time data further comprises real-time data relating to at least one of environmental conditions and driving conditions.