SFU Department of Statistics and Actuarial Science 2017 Undergraduate Case Study Competition:

Moj.io Car Sensor Data


The competition is posed by moj.io who has provided us with Real Data from car sensors.  Moj.io is hiring and wants to use this competition as a way to see what we can do.

Compete for $2000 in prizes and interviews with moj.io

Work in teams of up to 3

Eligibility: Any current SFU student at any level of their study;  BSc, MSc, PhD students are eligible from any department.  Maximum 1 PhD student per group.

Register here. Note that it can take up to 2 hours for you to gain access to the data link after signing up

Data Set: The data consists of over 2.4 million rows of sensor data from 384 cars and SUVs.

The data includes some information about the car type, engine size, transmission type, type of gasoline, factory advertised gas mileage,...

The sensors record variables like acceleration, engine revolutions per minute, velocity, fuel and battery levels, ...

Vehicle location information like latitude and longitude have been rounded to whole numbers.

Case Study Questions:

  1. 1)Produce a driver score.  Feel free to consider outside data sources like weather (See here for example).  This goal has its own prize!

  2. 2)GPS data is not included in the data, however, can you infer when someone is driving in the highway vs city?

  3. 3)Look at the differences in engine sensors to consider wear and tear on vehicles.  Produce a car score for wear and tear on vehicles.

4) Determine how long to drive to recharge the battery?

  1. 5)Does it seem like there might be multiple drivers per vehicle?   Find some cars that have these differences in behaviour.

Competition completes Monday April 24th

Questions?  email: Dr. Dave Campbell dac5@sfu.ca