Real-Time Driver Profiling & Risk Assessment for Usage-based Insurance with StreamAnalytix

1. Real-time Driver Profiling& Risk Assessmentfor Usage-based Insurancewith StreamAnalytix Tokeep upwith the new digital consumer andremaincompetitive, the auto insurance industry is increasingly investing inconnected car solutionstooffersimplified,transparent, andflexibleproductsandpricingoptions. For example, usage-based insurance is a voluntary, behavior-based insurance program that uses analytics to create highlypersonalizedanddynamicplansbasednotonlyon thedriver’sageandother demographics, butalsoaccounts forthedriver’sbehavior,risksrelated toavehicle, andexternal factors such asdrivingconditionsandweather.

2. About theCustomer This leading auto insurance provider chose StreamAnalytix to ingest, transform, enrich, analyze and store automotive telematicsdata inrealtimeto build anend-to-end analyticsapplicationfor driverprofiling&individual riskassessment, andsubsequentlyofferdynamic,usagebased,planstoitscustomers. Solution Highlights Real-time ingestion of telematics and sensordata The auto insurance company uses a telematics device to capture and transmit vehicle performance, usage, and driverbehaviordata fromvarious sensors inthecar.TheStreamAnalytix solutionenablesreal-time ingestion ofthis sensordata using anAWS(AmazonWebServices)IoTgateway.Thedevice captures datapointssuch as: Diverbehavior: Vehicle sensordata: Usagedata: Rapidacceleration, hard braking, hard cornering, and air bag deployment Oil temperature, engine performance, brake wear,andtire pressure mileage, location, andhistoric riskinessoftheroutesused Data processing, as itarrives In-memorydata transformation, data blendinganddata enrichmentisperformedasdriverbehavior, usage,and vehicle dataarrives: Combines real-time behavior and vehicle sensor data with riskhistory Blendsdrivingbehaviordatawith other real-time datasources such as syndicated public data marts and weatherdata Enriches data with customer information such as contact, location, age, pastpurchases, past claims, andmore Automated risk analysis through machinelearning Theingestion andenrichmentstagesprovidearich arrayof keyattributesneeded forthepredictivemachine learningmodelsrunningonApacheSpark. Orchestrated byStreamAnalytix, these stages assessandpredict individual riskscores. Classifying drivers as safe or risky and quantifying risk scores are based on current driving behavior, historical behavior, andsupplementaldata flowssuch as:usagedata,geographic location,vehicletype, vehicleperformance, andthird-partydata(like drivingconditionsandweatherdata). StreamAnalytix alsoprovideseasyvisualdev-ops interfaces toallowfor aperiodicrefreshofthemodelsbasedon varying patterns of data or drift in user behavior. If necessary and when appropriate, the insurer can configure StreamAnalytix todeployreal-time continuouslearningmodelssuchasK-meansclustering forthisusecase. www.streamanalytix.com

3. Smartalerting Theapplicationcreatesalertstoflagrisksbased onalteredbehaviorpatterns aswellasanomaliesinvehicle performance: Smart alert models are built to reduce false positives. For instance, if a driver is braking frequently because his route showsheavysnowfall, hisdriving behavior will not be flagged as a risk because there is a good explanation forit Customersmayopttobealerted in real-time on risks to enable course correction andcaution Alertsforvehiclehealthcanbe created toflagpredictedfaults and repair needs, reducingthe number of claims caused by vehiclebreakdowns Results An end-to-end, real-time analytics application for driver profiling & risk assessment to enable personalized, usage-based insuranceplans Through driver profiling and individual risk scores, the auto insurer is now offering data driven, and highly personalized,insurance policies andpricingplans. Additionally,theinsurance giantisnowalsooffering predictive maintenance services thatprevent vehiclebreakdowns andrepairs. Premiumadjustmentsanddynamicpricing Highlypersonalizedpremiumpricingoptionsarebasedon: Individualscores: Lowerinsurance premiumsforsafeorinfrequentdrivers Vehicle type andmake: Datashowsthatpeoplewithalowerriskprofileinherentlychoosecertain types ofcars Geography: Certain geographies havemore favorable weatherandbetter drivingconditions,leadingtolowerrisk and lowerpremiums www.streamanalytix.com

4. Increased customer loyalty and claims reduction from value-addedservices Remotevehiclediagnosticsandpredictivemaintenance services proved tobeconsumer-friendly value-additionwhich resulted in increased customer loyalty. Customers report that they like, and have come to rely on,v the application’s predictionsrelated tocomponentfailures andbreakdowns. Alsoincreasing preventative maintenance resulted in fewer claimsfromincidentscreatedby vehiclemalfunctions. Risk Distribution By Geography And VehicleMake Real-timetracking The application makes it easy to track driver activity and vehicle data in real-time through a custom web UI and interactive real-time dashboards. Customerscan alsoeasily tracktheirowndrivingbehaviorandvehicleperformance in real-time (through an installed mobile application) and take corrective action that can affect their insurance premiums. Real-time Dashboard for ActiveTrips www.streamanalytix.com

5. Connected Car Solution with StreamAnalytix Vehicle risk assessment Geographyrisk assessment Adjustedpremium prices Driver riskprofiling Real-time dashboards Predictive maintenance OtherApps Smart alerts End Device1 Alerts Driver behaviorand usage data Automobile Telematics device (Mqtt/http/ Web-sockets) End Device2 Smart car1 Vehicle performancedata AWS IOT Gateway (CentralAggregation Server/ Data flow Manager) Device provisioningand management (identity/ registrationetc.) Blending with datafrom open datamarts Storage andoffline analytics EndDevice 1 Automobile Telematics device (Mqtt/http/ Web-sockets) Data enrichment with historical and3rd party customerdata EndDevice 2 Machine learningand predictiveanalytics Smart car2 Alerts On-premise: Bare-Metal and/or VMs |Public/Hybrid Cloud Dataflow Controlflow © 2018 Impetus Technologies,Inc. All rights reserved. Product and company names mentionedhereinmaybetrademarks of their respectivecompanies. StreamAnalytix is an enterprise grade, visual, big data analytics platform for unified streaming and batch data processing based on best-of-breed technologies. It supports the end-to-end functionality of data ingestion, enrichment, machine learning, action triggers, and visualization. StreamAnalytix offers an intuitive drag-and-drop visual interface to build and operationalize big data applications five toten timesfaster,acrossindustries,dataformats,andusecases. open source Visit www.streamanalytix.com or write to us at inquiry@streamanalytix.com