About the Data

UBC's SkySpark IoT platform, managed by UBC Energy and Water Services (EWS), collects data on weather and UBC buildings every 15 minutes. The Urban Data Lab (UDL) is streaming the SkySpark data into an InfluxDB instance to increase its accessibility and usability.

InfluxDB 2.0 Instance

Public users (read permissions only) can log in to our InfluxDB 2.0 User Interface with the following credentials

• Username:public02
• Password: contact UDL ([email protected]) to request password

Public users can also access this InfluxDB instance from InfluxDB command line interface or InfluxDB API client libraries using an authorization token - Please contact UDL ([email protected]) to request a token.

The Python tutorial demonstrates querying the InfluxDB database using the influxdb-client Python module. Please contact UDL if you have any questions.

Database Structure

The SKYSPARK database is a hitorical copy of the SkySpark data in 2020, and the live streaming has been terminated. Please refer to the diagram below for its structure.

Currently, UDL is developing the database SKYSPARK2 which will provide cleaner data on more buildings.

InfluxDB concepts: database, point, measurement, tag, field, timestamp

1. For each data point on SkySpark, the id in POINT is the same as the uniqueID in READINGS.
2. A SkySpark data point can have up to 155 tags (recorded in the POINT measuremt). Each data point usually has the following tagging information so we also record these tags in READINGS to simplify queries.
   • siteRef: a site is a facility (usually a building)
   • groupRef: a group refers to a system or a floor in a building
   • typeRef: type of the data point
   • equipRed: information on the equipment
   • navName: information interpreted and added by SkySpark managers
   • unit: the unit of the reading value
3. Each data point has the reading values stored in one of the following fields in READINGS:
   • val_num: float values
   • val_bool: boolean values
   • val_str: string values
   • val_unk: value format unknown and stored as string
4. Timestamps are in UTC on InfluxDB: time in READINGS is the time the reading values are recorded, and time in POINT is a constant value to allow overwriting the existing data (this happens at 2am daily).