SparkT is a fully-featured, white-label modular analysis platform for driving style assessment, developed and owned by Sparkbit, a member of the PrediqTank consortium. SparkT recognizes dangerous driving-related events, such as tailgating, line hopping, speeding, red-light crossing. As such, it is a perfect monitoring and evaluation tool for traffic management engineers, fleet operators, and insurance companies.
Industry: Automotive, Telematics, InsurTech
Project Category: AI, data analytics, IoT, software development
Lead Member: Sparkbit
As a complete telematics solution, SparkT offers a multistep approach to driver behavior analysis:
SparkT collects data from multiple sources, including accelerometers, OBD2, GPS, and cameras.
The application detects hazardous events, such as speeding, excessive accelerations, or any dangerous driving habits associated with car control and traffic rules. Additionally, the application is capable of classifying weather, and recognizing road signs.
Using the collected data, SparkT builds a driver’s profile for future safety and for driving economy evaluation. The results and main statistics are visualized in the in-app dashboard, along with the driving style improvement tips.
SparkT can help not only individual drivers to improve their driving styles, but it can also be used by fleet managers or insurance companies, who can use the dashboards with advanced statistics to optimize fleet performance and evaluate their pilots.
Learn more at: https://sparkbitlabs.com/
Technologies used: Kotlin, Tensorflow, Keras