Mobile Sensor Readings of Activities

This collection contains mobile sensor readings used to train an activity recognition classifier. The sensor data was accessed via the Generic Sensor API. Mobile sensor readings from the following sensors:

• accelerometer
• gyroscope
• magnetometer
• gravity sensor
• relative orientation sensor
• absolute orientation sensor
• linear acceleration sensor

Recording were measured while the following activities were measured:

• Lying
• Sitting
• Standing
• Walking
• Stairs up
• Stairs down
• Running
• Taking the car
• Taking the bus
• Taking the tram
• Taking the train
• Taking the metro

Structure of the dataset

The collected data in each recording includes the start time timestamp readingID, the activity label activity, the duration
of the recording in elapsedTime, and an object sensorData, which contains the lengthy sensor measurements. The measurements for each sensor is stored in the form of an array containing on each index a series of four values. The time of when the measurement was taken t, and the value in the three axes x,y and z.

A minimalistic example of the format of the collected data:

{
"readingID": 1710151914418,
"activity": ""walk"",
"elapsedTime": 17312
"sensorData": {
"acce": [
...
{
"t": 1624,
"x": 0.05,
"y": -0.61,
"z": 0.21,
},
... ]
"gyro": Array (844)
"magnet": Array (136)
}
}

A big thanks to all the volunteers who helped collect this dataset.