Building Performance Analytics
A series of algorithms used to process real-time operational and historical data to provide intuitive analysis of system processes to identify trends and faults as well as other potentially useful data patterns to help, for example, with resource planning and energy exception reporting.
Essentially true Analytics are a set of rules (software) which are created and applied to a BMS to automatically monitor the performance of the system in simple terms. e.g. an experienced BMS operator analysing a dynamic AHU graphic may see differing temperatures on either side of a heating coil where the system is indicating a closed valve. He would determine that the valve may be 'letting by' and go to investigate. With Analytics applied the software would look at the relevant sensors and valve positions (requires comprehensive point count to work effectively) and automatically tell you the valve may be letting by. It could be argued that this therefore 'dumbs down' the BME Engineer's role but someone still has to act on the information.
Analytics require a whole different level of engineering in addition to the normal BMS functionality although a take up increases it is expected to get more standard rules and definitions 'out of the box'. However, factoring in the 'iBMS' concept any Analytics overlay would be bespoke and very complex and would need to be carefully considered.
It could reside both at Server and controller level as long as the device has the capability to process data from anywhere. For example, the Tridium JACE has that capability, a simple DDC controller could not.
As part of our 'Model-IB' design process the necessary Analytics are specified to determine and test the operation and value of any iBMS smart process and feature.