Click here to download the dataset.
Description of the Dataset
In our study, we collected the following data:
Smartwatch data:
- are located in the smartwatch.csv and smartwatch_interpolated.csv files
- smartwatch.csv:
- the different sensors are encoded into one column („SensorName“)
- 6 sensors: linear acceleration, gyroscope, gravity, game rotation vector, accelerometer and heart rate sensor
- left and right hand are coded into one column separately („Sensor on Hand“)
- no processing of the data
- this file contains the raw data of the smartwatch sensors, which has been gathered and merged into one file
- smartwatch_interpolated.csv:
- difference to previous file: sensors are encoded into different columns, as well as their corresponding hand (e.g. „Y_linearAccelerationSensor_right“, „Z_gravitySensor_left“ etc)
- unfortunately the six sensors provide data with different sampling rates, which leads to different timestamps and many empty cells in the data
- the resulting empty cells have been filled by interpolation with 100ms
OptiTrack data:
- located in optiTrack.csv
- contains rotation and position data of both hands and the LEGO plane
- the data has been sampled with 120Hz
- no processing of the data
- we gathered the different raw sensor data files and put them together into one file
LEGO data:
- located in lego_data.csv
- contains both Smartwatch and OptiTrack data
- we merged the columns of smartwatch_interpolated.csv and optiTrack.csv file
- we synchronized the timestamps and filled the empty cells by an interpolation of 100ms
In addition to the sensor data for all files, we also added the
- Participant ID (“Participant),
- Task ID (“Task”)
- Variable I: paper-based instruction (not) available ( 1 = yes, 0 = no)
- Variable T: (no) time limit ( 1 = yes, 0 = no)
- Variable B: number of building bricks ( 0 = normal, 1 = less, 2 = more)
- Dominant Hand: participants’ dominant hand ( 1 = right, 0 = left)
- Watch Hand: hand on which participants normally wear their watch ( 1 = right, 0 = left)