Data Storage

JTrack stores smartphone, wearable, and EMA data using a structured, BIDS-inspired organization. Data is grouped by study, participant, and data source, facilitating longitudinal analyses and reproducible research.

Data Sources

The platform collects three primary categories of data:

• Smartphone sensor data

• EMA questionnaire responses

• Wearable device data (e.g., Garmin)

Each data source is stored independently and linked through the participant identifier.

Smartphone Sensor Data

Smartphone sensors generate timestamped JSON records. Each sensor type produces its own data structure depending on the information collected.

Example: Activity Recognition

{
     "activityType": "WALKING",
     "confidence": 100.0,
     "deviceid": "fa95d3-9d74-4c4f-8def-8bee65276",
     "sensorname": "activity",
     "studyId": "Demo_Study",
     "timestamp": 1690969600686,
     "username": "Demo_Study_00001_1"
 }

Example: Application Usage

{
     "appName": "Quickstep",
     "beginTimeStamp": 1690969205734,
     "endTimeStamp": 1690969486927,
     "lastTimeUsed": 1690969262708,
     "totalTimeinForeground": 32446,
     "deviceid": "fa62d3-9d74-4c4f-8def-8be6e65276",
     "sensorname": "application_usage",
     "studyId": "Demo_Study",
     "timestamp": 1690969606410,
     "username": "Demo_Study_00001_1"
 }

Example: Location

{
     "accuracy": 20.0,
     "alt": 32536.115439935587,
     "lat": 50.5519819674761,
     "lon": 6.319306026624161,
     "provider": "fused",
     "deviceid": "fa562d3-9d74-4c4f-8def-8b3b665276",
     "sensorname": "location",
     "studyId": "Demo_Study",
     "timestamp": 1690969600504,
     "username": "Demo_Study_00001_1"
 }

EMA Data

EMA responses are stored as timestamped question-answer records.

Example:

{
     "username": "Demo_Study_00001_1",
     "studyId": "Demo_Study",
     "subjectId": "",
     "deviceid": "f562d3-9d74-4c4f-8def-8b365276",
     "questionId": 1,
     "answerId": 1,
     "timestamp": "1778938864409",
     "answerText": null,
     "answer": null,
     "freeText": "Sehr beeinträchtigt",
     "sensorname": "ema",
     "startTime": "1778938782793",
     "endTime": "1778938781362",
     "surveyVersion": 12,
     "databaseId": 44232
 }

Garmin Wearable Data

Garmin devices can provide physiological and activity-related measurements, depending on the device model and enabled sensors.

Example: Heart Rate

{
     "deviceName": "vívoactive 6",
     "deviceid": "aaac108-2886-4d1c-85ed-8f154ecf22",
     "frequency": "15m",
     "id": 9,
     "studyId": "Demo_Study",
     "timestamp": 1779139800000,
     "timestamp_end": 1779140699999,
     "timestamp_start": 1779139800000,
     "username": "Demo_Study_00001_1",
     "value": 81.0,
     "sensorname": "garmin",
     "wearable_sensor": "HEART_RATE"
 }

Example: Respiration

{
     "deviceName": "vívoactive 6",
     "deviceid": "aac108-2886-4d1c-85ed-870f4ecf22",
     "frequency": "15m",
     "id": 12,
     "studyId": "Demo_Study",
     "timestamp": 1779139940000,
     "timestamp_end": 1779140839999,
     "timestamp_start": 1779139940000,
     "username": "Demo_Study_00001_1",
     "value": 13.0,
     "sensorname": "garmin",
     "wearable_sensor": "RESPIRATION"
 }

Example: Steps

{
     "deviceName": "Forerunner 255",
     "deviceid": "520dd6-2889-4385-bf-9f15f1214001",
     "frequency": "15m",
     "id": 38,
     "studyId": "Demo_Study",
     "timestamp": 1779213600000,
     "timestamp_end": 1779214499999,
     "timestamp_start": 1779213600000,
     "username": "Demo_Study_00001_1",
     "value": 24.0,
     "sensorname": "garmin",
     "wearable_sensor": "STEPS"
 }

Data Export

JDash provides functionality to export both raw and processed study data.

Raw Data Export

Raw exports contain the original JSON records collected from smartphones, wearable devices, and EMA questionnaires. These exports preserve the original data structure and timestamps as recorded by the JTrack applications.

Examples include:

• Smartphone sensor records

• EMA questionnaire responses

• Garmin wearable measurements

Processed Data Export

Processed exports transform raw records into tabular CSV files suitable for quality control, reporting, and downstream analysis.

Currently available exports include:

• Data-per-day summaries per subject

Additional processed export formats will be added in future releases.

Note

Processed CSV files are generated from the underlying raw JSON records and may include derived metrics, aggregations, or quality-control indicators depending on the selected export type.