Model : Driving Behavior Classification



Required Data

  • Driving Characteristics : Speed, Acceleration

  • Meteorology Data : Weather, Temperature

  • Traffic Conditions : Day of the week, Time

  • Road Characteristics : Road type, Bus stops


Application Scenarios

  • For Indivisual : Bihavior based vechicle managemant, Energy efficient routing

  • For Company : Optimal Allocation of Charging Network

  • For Government : Energy load forecasting(time, location)


How to create hyper-personalized AI model of 'Driving Behavior Classification'

  • The mobility domain can be divided into three key areas: vehicle diagnostics, in-cabin situations, and V2X/G communications.

  • Each category involves handling highly sensitive personal information, such as location data, conversations, facial recognition, and authentication.

  • By applying CANDiY AI technology, which enables AI models to learn directly on the user’s device, we can minimize the risks associated with processing sensitive data.


Use Case : Ongoing PoC with an Energy Company in the Netherlands. (Aug.'24 ~ Nov.'24)