Cambridge Mobile Telematics Inc.

William Francis Bradley

Hari Balakrishnan

Jun-geun Park

Samuel Ross Madden

Kimberly Shea

Geronimo Mirano

Lewis David Girod

Greg Padowski

Paresh Govind Malalur

Katherine Esther Dalis

Gregory David Padowski

Juan Martin Munoz Valdez

A method for constructing and evaluating a statistical model includes receiving, by a data processing system, telematics data and insurance claims data for a population of drivers. A training dataset is generated based on the telematics data that includes values for a proxy variable derived from the telematics data, and values for one or more features derived from the telematics data. A testing dataset is generate based on the telematics data and the claims data that includes values for a target variable derived from the claims data, and values for the one or more features derived from the telematics data. A statistical model is generated using the training dataset, the statistical model configured to predict values of the proxy variable from values of the one or more features. The statistical model is validated using the testing dataset.

Constructing a statistical model and evaluating model performance

A system can include a mobile device that includes one or more sensors for sensing information during a trip in a vehicle. A hardware processor can execute operations including receiving telematics information produced by one or more sensors during a trip in a vehicle; processing, by a hardware processor, the received telematics information to identify vehicle movement information for the vehicle during the trip; determining, by the hardware processor, a probability that an advanced driver assistance system (ADAS) feature of the vehicle was operational during the trip based, at least in part, on the vehicle movement information for the vehicle during the trip; and determining, by the hardware processor, a risk score for the vehicle or a driver of the vehicle based, at least in part, on the probability that the ADAS feature was operational.

Detecting use of driver assistance systems

Among other things, a documentation of a crash involving a vehicle is generated automatically. Telematics data is received that has been produced by one or more sensors associated with a telematics device at the vehicle. Based on the telematics data, a vehicle crash period is determined that begins at a start time and ends at an end time of the vehicle crash. Based on the telematics data, one or more metrics are determined associated with the vehicle during the vehicle crash period. Based on one or more metrics, a human-readable documentation of the vehicle crash is generated automatically.

Vehicle telematics of vehicle crashes

In an aspect, a fuel economy of a vehicle is determined by: receiving telematics data produced at least partly by one or more sensors at a vehicle; identifying, based at least partly on the telematics data, one or more fueling events of the vehicle; determining, for each fueling event of the one or more fueling events, an amount of fuel added to the vehicle for that fueling event; and determining, based at least partly on the amount of fuel added to the vehicle for each fueling event and the telematics data, a fuel economy of the vehicle.

Estimating fuel economy

Among other things, safety performance of vehicles is monitored or evaluated, or both. Sensor data is received from one or more sensors at a vehicle. One or more risk sources encountered by the vehicle during a period of operation of the vehicle are identified based on the sensor data. One or more safety events experienced by the vehicle during the period of operation are identified based on the sensor data. A risk of a safety event experienced by the vehicle during the period of operation is determined based on the one or more risk sources. A safety performance of the vehicle during the period of the vehicle during the period of operation of the vehicle is determined based on the one or more safety events and the risk.

Evaluating the safety performance of vehicles

Among other things, instructions stored in a tangible storage medium are executable by a processor to receive data representing an attitude of a vehicle being driven relative to a frame of reference and process the received attitude data to determine a risk factor in the driving of the vehicle.

Reducing driving risk

Among other things, information generated by sensors of a mobile phone and indicative of motion of the mobile phone and state information indicative of a state of operation of the mobile phone are monitored. Based on the monitoring, distraction by a user of the mobile phone who is a driver of a vehicle is determined.

Determining, scoring, and reporting mobile phone distraction of a driver

An approach to determining vehicle usage makes use of a sensor that provides a vibration signal associated with the vehicle, and that vibration signal is used to infer usage. Usage can include distance traveled, optionally associated with particular ranges of speed or road type. In a calibration phase, auxiliary measurements, for instance based on GPS signals, are used to determine a relationship between the vibration signal and usage. In a monitoring phase, the determined relationship is used to infer usage from the vibration signal.

Mileage and speed estimation

Customize safe speeds for vehicles are determined from information that is collected from multiple trips and multiple drivers across many road segments. The information collected includes the speed traveled along the road segment, possibly along with additional information, such as (for example) driver, time of day, day of week, weather and road conditions, angle of the sun or other factors. Given information from a new trip, the driver's speed can be compared on each road segment with relevant data from previous trips (from e.g., similar roads, weather conditions, drivers). The comparison yields a score representing the safety of the driver's speed.

Growing community of inventors

Arlington, MA, United States of America

William Francis Bradley