sMAP: Simple Measurement and Actuation for Integrating and Managing Physical Data

Time-series data is increasingly ubiquitous and generated in large volumes; commercial buildings are a particularly rich source of this information generated by the building management system, computer networks, occupant activity, and many other systems. When put to use, this information could lead to significant reductions in energy consumption; however, it is often located in proprietary "silos", fragmentary, and otherwise disorganized. The sMAP project aims to create a simple, common foundation for collecting and organizing this data: in this talk, we discuss the challenges of dealing with the heterogeneous data sources and providing a simple platform for building energy application. We detail some of the tens of thousands of streams currently collected, and demonstrate some of the many uses sMAP has been put to.

Speaker: Stephen Dawson-Haggerty, UC Berkeley