Systems and methods for in-vehicle predictive failure detection

1. A method, comprising:

monitoring one or more ECUs of a vehicle to detect diagnostic trouble codes (DTCs);

comparing one or more detected DTCs to one or more rules or trained model objects;

determining a probability of failure of the vehicle based on the one or more detected DTCs, the probability of failure being supplied by the one or more rules or trained model objects;

for a training set including a plurality of sessions, deriving a threshold probability by:

identifying a set of all possible patterns of DTCs for the plurality of sessions,

estimating a probability of failure, using Bayes' theorem, for each possible pattern of DTCs of the set of all possible patterns of DTCs,

creating an ordered list of possible patterns of DTCs by ordering the set of all possible patterns of DTCs by the probability of failure for each possible pattern of DTCs,

determining a sensitivity curve and a specificity curve for the ordered list of possible patterns of DTCs, and

selecting the threshold probability at an intersection of the sensitivity curve and the specificity curve; and

displaying one or more instructions for an operator of the vehicle via one or more output devices of the vehicle in response to the probability of failure for the pattern of one or more generated DTCs exceeding the threshold probability,

wherein the probability of failure of the vehicle is dependent upon a sequential order of detection of the one or more detected DTCs.

2. The method of claim 1 , wherein the trained model objects are generated using machine learning algorithms performed on historical DTC data.

3. The method of claim 1 , wherein the determining is further based on a plurality of operating conditions comprising an odometer reading and a battery voltage.

4. The method of claim 1 , wherein the one or more instructions include a textual message displayed via a screen.

5. The method of claim 4 , wherein the instructions include a recommended number of days in which to visit a service station.

6. The method of claim 5 , wherein the recommended number of days is based on the probability of failure for the one or more detected DTCs, a greater number of days being recommended for a lower probability of failure and a smaller number of days being recommended for a higher probability of failure; and wherein the number of days is further based on a vehicle subsystem generating the DTCs.

7. The method of claim 1 , wherein the probability of failure of the vehicle is dependent upon a mileage of the vehicle.

8. The method of claim 1 , wherein the probability of failure of the vehicle is dependent upon an engine load of an engine of the vehicle.

9. The method of claim 1 , wherein the probability of failure of the vehicle is dependent upon a speed of the vehicle.

10. The method of claim 1 , wherein the one or more instructions for the operator of the vehicle depend upon the sequential order of detection of the one or more detected DTCs.