Data Schema of the Building Energy Metering Network

The Data Schema specifies the data structure that will be manipulated by the Data Processing Architecture. More specifically, it describes how the entities of the Data Acquisition domain relate each other, in order to efficiently execute queries over the data of this domain. The main entities of the schema are as follows:

• Location. Is the building location that is being monitored by a given Device. Which, in the scope of this work, always assumes the form of an Energy Meter. Moreover, the entity is described through its building location and covered area.
• Device. Is the entity that represents all types of sensors deployed at the building, such as: Energy Meters, Equipment Sensors, and Environment Sensors. As said before, in this work, it serves to represent the Energy Meters.
• Datapoint. Identifies the measurement points composing each device. In our case, each device (energy meter) has three datapoints, one to measure each phase of the tree-phase current that power supplies the building. The entity is described by the description and the unit of its measured value.
• Datapoint Reading. Are the time-variant values periodically produced by the datapoints of each device. In the case of this work, given the existence of three datapoints per device (three-phase current), each device's produced measurement is composed by three data stream tuples. Each one of the three datapoint readings is described by a timestamp, that will be the same for the three tuples belonging to the same measurement, and by the tuple’s value of each current-phase.

The manner how the data schema is addressed depends on the type of query engine that supports the Data Processing Architecture. Being the different approaches of addressing this data schema strongly related with the main concern of this work:

• In the DBMS Based Solution, all six entities are persisted in the database. This is the most suitable approach for the five static data relations. However, streaming data relation Datapoint Reading will also be persisted, meaning that the energy metering data streams will be processed offline.

• In the DSMS Based Solution, just the five static data relations are persisted in the database. Whereas the relation Datapoint Reading is processed on-the-fly by the DSMS, instead of being persisted in the database. Meaning that the energy metering data streams will be processed online.

---

Appendix of the conference paper: "Real-Time Monitoring of Building Energy Metering Networks", (Anjos et al., 2015).