Data Insights from Vehicle Time Series Messages

Data Insights from Vehicle Time Series Messages

Mentor: Borhan Sanandaji

Summary

  • Idea 1: Characterize and visualize the make/model/year distribution in the dataset; extend this by characterizing the conditional distribution over the geographical span of the dataset.

  • Idea 2: Determinine the devices in the dataset that have multiple drivers per mojio device

  • Idea 3: (Optional) Using any of the existing variables in the dataset or outside resources, forecast the battery voltage of the vehicles.

About the data

~950k device messages each including 158 variables of 166 different cars sampled over 3 week time period.