Telematics - Papers

Datasets

• levin vehicle telematics
• Automobile Telematics Dataset
• Safe Driver Telematics
• Example Driving Data (telematics)
• Synthetic Dataset

Paper - Synthetic Dataset Generation of DriverTelematics

Features

Further Datasets - Link

Paper - Collaborative Cloud-Edge Computation for Personalized Driving Behavior Modeling

Key Notes

• Generative Adversarial Recurrent Neural Networks (GARNN)
• CGARNN-Edge (Conditional GARNN)
• Driving behavior modeling can also be used by insurance companies to determine the vehicle insurance premium
• Drivers may have distinct driving behaviors because of their individual difference, such as age group, gender, and driving experience
• Real-time performance is a stringent requirement for ADAS. For example, fatigue driving or other abnormal driving behaviors should be detected immediately

• Driving behavior: speed, acceleration, brake force, steering, lane offset, and lane position signal

Paper - Driver Telematics Analysis

Key Notes

• Aggressive behaviors include lane violations, failure to stop, speeding, sudden raise of acceleration and severe other violations
• Behavior parameters that account for aggregate driving profile: mean speed, mean speed excluding the stops, mean acceleration, mean deceleration, average length of a trip, mean number of acceleration/deceleration changes within a trip, standstill time proportion, acceleration time proportion, deceleration time proportion and constant speed time proportion
• Trip Features - Ride Length, Ride Speed, Ride Length without stops, Ratio of Stops, speeds, angles, accelerations, speed*angles.
• Driving features - Mean acceleration, Mean deceleration, Average number of acceleration/deceleration changes

Paper - Analyzing driving behavior from CAN data using context-specific information 

Key Notes

Speed based acceleration thresholds - applicable to all categories

• Here adaptive speed - based thresholding is derived exponential regression equations
• Thresholds for turn / straight segments are different. Stricter for turns

Paper - Driving Style Representation in Convolutional Recurrent Neural Network Model of Driver Identification∗

Key Notes

• Input: A trajectory 𝑇 .
• Model: A predictive model 𝑀 to capture variations in driving behavior to derive driving style information.
• Goal: Predict identity of driver for trajectory 𝑇 based on driving style information.
• Optimization Objective: Minimize prediction error

A Vehicle Classification Algorithm based on Telematics Data

Keep Exploring!!!

Driver Profiling - Telematics - Connected Cars

Sometimes we need more data pointers, research paper reads for a use case. arxiv / phd papers are a wealth of information to get right directions for dataset, features, models

Paper - Telematics and Contextual Data Analysis and Driving Risk Prediction

Key Notes

• Usecase - Driving Risk Prediction
• Solution that consists of three parts
• a) characterizing driving context, 
• b) characterizing driving style, and 
• c) context-aware driving risk prediction

Areas

• Driving context can be described as a combination of location (e.g., Interstate-90) and time (e.g., weekdays between 3pm to 7pm).
• Characteristics of contexts from the aggregate behavior of drivers Segmenting trajectories to identify meaningful driving patterns
• Analyze each pattern with respect to contextual data to identify cause-andeffect patterns of significance

Features

• Contextual data - traffic data
• Weather data - weather stations
• Road-network characteristics - (e.g., road type and road shape)

Analysis of telematics alongside contextual data provides valuable insights regarding an individual’s behavior, common driving habits, and characteristics of the road network with regard to dynamic traffic flow

• Roads Analysis - sharp-turn , smooth-turn, exit/merge, intersection, exit/merge, ramp, bridge
• Drive Analysis - speed, acceleration, GPS coordinates, heading
• Time features - Type of Day, Time of the Day, frequency of congestion events
• For traffic data, we have loc = (latitude, longitude, Street Name, Street Side, Zipcode, City, State)

Patterns

• Monthly Traffic Distribution
• Monthly Weather Distribution
• Weekly Traffic Distribution

Weather Entity

• Severe-Cold: the case of having extremely low temperature, with temperature 
• Fog: the case where there is low visibility condition as result of fog or haze.
• Hail: the case of having solid precipitation including ice pallets and hail.
• Rain: the case of having rain, including any type of the rain, ranging from light to heavy.
• Snow: the case of having snow, including any type, ranging from light to heavy.
• Storm: the extremely windy condition, where the wind speed is at least 60kmh.
• Precipitation: a generic label which we frequently observed in raw weather data, however, we have no further information to include them in any of the previously described entity types

Road-network - Interstates and Freeways, Cities

Clustering based on Traffic patterns

Study the behavior of an individual driver in order to evaluate how risky or safe he/she is

Common propagation patterns of traffic and weather entities

rain → accident → congestion

major construction → more congestions

Tree-pattern-mining-based process, which we name short-term pattern discovery

Input: A trajectory T.

Model: A predictive model M to capture variations in driving behavior to derive driving style information

gps, accelerometer, and magnetometer

Contextual  data such as traffic events, weather data, points-of-interest, and time

Comprehensive set of attributes to describe each accident including location data, time data, natural language description of event, weather data, period-of-day information22, and relevant points-of-interest data

Dataset - US Accidents

Dataset which is called the 100-car naturalistic driving study

Cluster based on accident history leveraging accident datasets

Using telematics data alone for driving risk prediction is a recent trend given notable attention in the past few years

Risk assessment for individual drivers, based on crash and near-crash

• (CNC) events, as well as critical-incident events (CIE), age, and personality of drivers to be the important risk factors 
• Using K-Means clustering, they performed clustering of CNC rates, and identified three clusters of low, moderate, and high risk drivers
• Average monthly drive time, age, gender, living region, and car’s age, to predict the frequency of claims for different drivers
• Coarse-grained attributes such as yearly distance, number of trips, average Distance per trip, and coverage of different road types by distance
• Driving state variables (e.g., sharp-turn, lane-change, abnormal acceleration/deceleration, and speeding with respect to speed-limit data)
• (a) Smooth turn (b) Sharp and smooth turn (c) Sharp turn trajectories

Very Good Phd Document :)

Loved it, A lot of inspiration to apply it in use cases.

More reads

• Discovery of Driving Patterns by Trajectory Segmentation
• Artificial Intelligence and Data Science in the Automotive Industry
• Synthetic Dataset Generation of Driver Telematics
• SURVEYING OFF-BOARD AND EXTRA-VEHICULAR MONITORING AND PROGRESS TOWARDS PERVASIVE DIAGNOSTICS
• Emerging Autonomous Vehicle Risks: The Role of Telematics-Based Risk Assessment
• Predictive Maintenance for Industrial IoT of Vehicle Fleets using Hierarchical Modified Fuzzy Support Vector Machine

Happy Reading!!!

Connected Cars - Telematics Use cases - Research Paper

Paper - How much telematics information do insurers need for claim classification?

Key Notes

• By comparing a few classification algorithms, we find that logistic regression with lasso penalty is the most suitable for our problem
• Telematics data become redundant after about 3 months or 4,000 kilometers of observation, at least from a claim classification perspective
• Telematics data fall under the definition of “personal data”, and must therefore be handled according to the relevant legislation
• In this analysis, only collision coverage claims were considered, i.e. the target column given as input to the classification models is the indicator of a collision claim, at-fault or not
• Instead of having the indicator of a claim as the response variable, we would instead have the number of claims, moving us into a counting regression context
• Low claims / Medium / High claims we could bucket and assign premium accordingly

Classical features selected for the analysis

“usage”, “travel habits” and “driving performance”. Claimants (those who have claimed at least once during their observed year) and the non-claimants (those who have not claimed during their observed year)

Summary of Features

Collected Feature

• VIN
• Trip number
• Departure datetime 
• Arrival datetime 
• Distance 
• Maximum speed

Derived Features

• annual_distance
• commute_distance
• conv_count_3_yrs_minor
• gender
• marital_status
• pmt_plan
• veh_age
• veh_use
• years_claim_free
• years_licensed

Derived Features by “usage”, “travel habits” and “driving performance”

• avg_daily_distance
• avg_daily_nb_trips
• med_trip_avg_speed
• med_trip_distance
• med_trip_max_speed
• max_trip_max_speed
• prop_long_trip
• frac_expo_night
• frac_expo_noon
• frac_expo_evening
• frac_expo_peak_morning
• frac_expo_peak_evening
• frac_expo_mon_to_thu
• frac_expo_fri_sat

More Reads

• Evolution of Insurance: A Telematics-Based Personal Auto Insurance Study
• Driver Identification Based on Vehicle Telematics Data using LSTM-Recurrent Neural Network
• Vehicle Telematics Via Exteroceptive Sensors: A Survey
• Secure OTA Software Updates in Connected Vehicles: A survey

Keep Thinking!!!