Use accelerometer and gyroscope data from smartphones to identify vehicle type (bus or car) and phone location (driver side or passenger side).
- Extract gravity signal (see "Extract Gravity Signal" jupyter notebook)
- Rotate XYZ signals to vehicle reference frame (see "Rotate Sensor Data to Vehicle Reference Frame" jupyter notebook)
- Resample time series data to 10 Hz sampling rate (see "Resample Sensor Data to 10 Hz Sampling Rate" jupyter notebook)
- Automate steps 1-3, modify columns to standard system, and save result to a new file (see "Process Smartphone Sensor Data" jupyter notebook)
Two examples of use cases of this software:
- Vehicle classification exercise: determine whether the smartphone is on a bus or in a car based on 5-10 minutes of sensor data (see "Vehicle Classification Exercise" jupyter notebook)
- Phone position classification exercise: determine whether the smartphone is on the driver side or passenger side of a car based on 5-10 minutes of sensor data (see "Phone Position Classification Exercise" jupyter notebook)