SYSTEMS AND METHODS FOR IDENTIFYING DISTRACTED DRIVING EVENTS USING UNSUPERVISED CLUSTERING
A distracted driving analysis system for identifying distracted driving events is provided. The system includes a processor in communication with a memory device programmed to: (i) receive driving event records including phone usage by a user that occurred within a time period of a driving event, (ii) divide the driving event records into at least two clusters based at least in part upon common features of one or more driving event records of the plurality of driving event records by processing the driving event records using an unsupervised machine learning algorithm, (iii) generate a trained model based at least in part upon the at least two clusters including cluster labels, (iv) process a new driving event using the trained model, (v) assign the new driving event to one of the at least two clusters using the trained model, and (vi) based at least in part upon the cluster labels for the assigned cluster, determine whether the new driving event is an actual distracted driving event or a passenger event.
1 - 20 . (canceled)
21 . A system for identifying distracted driving events comprising:
a distracted driving analysis device including one or more processors programmed to provide an event module, a machine learning module and a trip analysis module;
wherein the event module is configured to process driving event records to determine potential distracted driving events;
wherein the machine learning module is configured to receive features input from a user computing device to cluster the potential distracted driving events determined by the event module into clustered data by processing the potential distracted driving events using a clustering algorithm based at least in part upon features defined in the features input;
wherein the machine learning module is further configured to receive clusters qualifications from the user computing device qualifying the clustered data as either vehicle operator clusters or passenger clusters and define a trained model based at least in part upon the qualified clustered data; and
wherein the trip analysis module is configured to analyze user data using the trained model to identify distracted driving events and passenger events represented by the user data.
22 . The system of claim 21 , wherein the driving event records include historical driving data associated with operation of a vehicle and historical phone usage data associated with usage of a mobile computing device.
23 . The system of claim 22 , wherein the historical driving data includes at least one of data collected by a vehicle sensor, data collected by a vehicle operating system, or data collected by the mobile computing device.
24 . The system of claim 22 , wherein the historical phone usage data includes at least one of application-related data, texting data, or general phone usage data.
25 . The system of claim 22 , wherein the event module determines potential distracted driving events based at least in part upon timestamps within the historical driving data and the historical phone usage data.
26 . The system of claim 21 , wherein the features input includes at least one of acceleration data, speedometer data, braking data, tap and swipe data, or texting data.
27 . The system of claim 21 , wherein the features input includes compound features including at least one of a combination of two or more data types or a relationship between two or more data types.
28 . The system of claim 21 , wherein the cluster qualifications include manual input denoting whether each clustered data likely indicates a distracted driving event or a passenger event.
29 . The system of claim 21 , wherein the clustering algorithm is a semi-supervised machine learning algorithm.
30 . The system of claim 21 , wherein the trip analysis module is further configured to determine a confidence level that the user data indicates a distracted driving event or a passenger event.
31 . The system of claim 21 , wherein the one or more processors is further programmed to provide a profile module that receives identified distracted driving events and passenger events from the trip analysis module and generates a driver profile of a user associated with the user data.
32 . A computer-implemented method for identifying distracted driving events comprising:
processing driving event records to determine potential distracted driving events;
receiving features input from a user computing device to cluster the potential distracted driving events into clustered data by processing the potential distracted driving events using a clustering algorithm based at least in part upon features defined in the features input;
receiving clusters qualifications from the user computing device qualifying the clustered data as either vehicle operator clusters or passenger clusters;
defining a trained model based at least in part upon the qualified clustered data; and
analyzing user data using the trained model to identify distracted driving events and passenger events represented by the user data.
33 . The computer-implemented method of claim 32 , wherein the driving event records include historical driving data associated with operation of a vehicle and historical phone usage data associated with usage of a mobile computing device.
34 . The computer-implemented method of claim 33 , wherein processing driving event records to determine potential distracted driving events includes evaluating timestamps within the historical driving data and the historical phone usage data.
35 . The computer-implemented method of claim 32 , wherein the features input includes at least one of acceleration data, speedometer data, braking data, tap and swipe data, or texting data.
36 . The computer-implemented method of claim 32 , wherein the cluster qualifications include manual input denoting whether each clustered data likely indicates a distracted driving event or a passenger event.
37 . The computer-implemented method of claim 32 , wherein the clustering algorithm is a semi-supervised machine learning algorithm.
38 . The computer-implemented method of claim 32 , further comprising generating a driver profile of a user associated with the user data.
39 . A system for identifying distracted driving events comprising:
a distracted driving analysis device including one or more processors programmed to provide a machine learning module and a trip analysis module;
wherein the machine learning module is configured to:
receive labeled training data indicating a distracted driving event wherein usage of a mobile computing device is conducted by a user who is operating a vehicle or a passenger event wherein usage of the mobile computing device is conducted by a passenger in the vehicle;
analyze the labeled training data to identify features that are correlated with a label of a distracted driving event and a label of a passenger event; and
train a model based at least in part upon the identified features; and
wherein the trip analysis module is configured to analyze user data using the trained model to identify distracted driving events and passenger events represented by the user data.
40 . The system of claim 39 , wherein the trip analysis module is further configured to determine a confidence level that the user data indicates a distracted driving event or a passenger event.
41 . The system of claim 39 , wherein the one or more processors is further programmed to provide a profile module that receives identified distracted driving events and passenger events from the trip analysis module and generates a driver profile of a user associated with the user data.
42 . A system for identifying distracted driving events comprising:
means for processing driving event records to determine potential distracted driving events;
means for receiving features input from a user computing device;
means for clustering the potential distracted driving events into clustered data;
means for receiving clusters qualifications from the user computing device qualifying the clustered data as either vehicle operator clusters or passenger clusters;
means for defining a trained model based at least in part upon the qualified clustered data; and
means for analyzing user data using the trained model to identify distracted driving events and passenger events represented by the user data.