Building Data Analytics

Transcription

Building Data Analytics
Building Data Analytics
Integrating building data with
critical analytics and interfaces
for a high performance built environment
Azizan Aziz
Assistant Research Professor
Carnegie Mellon University
Center for Building Performance and Diagnostics
What are the steps?
1. Collect, structure and store monthly and real time data
2. Automate BAS and other monitoring and diagnostics
3. Make accessible to Corporate, Campus and City Executives (ID-C)
4. Make accessible to Facility Managers (ID-F)
5. Make accessible to Occupants for control (ID-O)
6. Make accessible to the Public (Disclosure) & Research Community
Diverse Inputs + OSIsoft PI System + CMU Analytics + Microsoft BI 365/Azure ML + CMU User Interfaces
Azure SaaS
SQL 2012
Other
Sources
PI Cloud Connect
End User
Executive
(Office Occupants)
PI
ProcessBook
PC
Asset Framework
Server
Building
Attributes
Power BI for
Office 365
Energy
Bill
PI Interface
iPad
(Coresight Apps)
PI Server
Mobile
Submetering
Data
BAS
Siemens
JC
AL
PI Coresight
PI Coresight
Web Server
Custom
WebSites
CMU Dashboards
MySQL Database
Plugwise Source
Web Server
Individual
Portfolio Manager
ID-C
Intelligent Data for Corporate and Campus
Portfolio Managers
With Monthly Electric, Gas and Water Bills
Partners: PNC
CMU Center for Building Performance & Diagnostics
Managing a Building Portfolio
• Energy Manager at a Pittsburgh Bank
• Challenge: Monitor and analyze a portfolio of 4000+ buildings from
Headquarters to Data Centers, down to ATMs.
• With real-time utilities and building automation = 10 billion data points/year
Unstructured Data Model
Structured Data Model
Commercial Partner : Pittsburgh Bank
Query the Energy consumption breakdown of any building in 3 clicks
Commercial Partner : Pittsburgh Bank
ID-C Challenge :
To monitor and analyze a
portfolio of 4000 Assets
(Headquarters, offices,
datacenters, branches,
ATMs) for strategic
improvements
Engaging Stakeholders in Building Energy and Water Reduction
Kai-Wei Hsu & Ting Wang
Engaging Stakeholders in Building Energy and Water Reduction
Kai-Wei Hsu & Ting Wang
Engaging Stakeholders in Building Energy and Water Reduction
Kai-Wei Hsu & Ting Wang
Monthly utility consumption: Lean Analysis
0
ID-F
Intelligent Data for Facility Managers
With Real Time Building Automation Data and
Electric, Gas and Water Consumption
Partners: OSIsoft, DOE, DOD
CMU Center for Building Performance & Diagnostics
Building 101 Public Web-interface
http://128.118.67.245:85/eebhub/index.php
Electricity Consumption Trends by End Use and Summaries
Gas Consumption Trends by End Use and Summaries
Real-Time IEQ Status
Energy Map (8760 hours/1 Year)
Facility Manager Space and System Read-outs
More than 20 Universities can access and share real time building information
Two data plots, osisoft data base ID-F
Two data plots, osisoft data base ID-F
Online webpage and iPad interface
iPhone interface
PI Coresight iPad and iPhone interfaces for real-time and
trended data for Facility Management field work
Carnegie Mellon University
Pittsburgh Pennsylvania
Founded in 1900 by
Andrew Carnegie
12,991 Students
(6,223 undergraduate)
CMU annual energy
budget over $20M
That’s over $1,600 per
year per student!
Goal:
CMU to be a leading
university in
sustainability
About 6,500,000 sqft
65 + Buildings
80,000 data points
22
Temperature and Energy Prediction
1. Collect Data Real-Time
2. Train Model
3. Predict temperature and energy at
different time intervals (up to 48h)
4. Detect potential energy savings
• Over-Cooling/Heating
• Space conditioned without
occupancy
5. Corrective Actions:
• Adjust Control Logics
• Turn Off systems
OSIsoft PI Coresight
23
Thank You
Azizan Aziz
[azizan@cmu.edu]
Project Funded by
Department of Energy [CBEI]
Department of Defense [ESTCP]
Demo
• 128.2.109.227/Coresight
• 128.118.67.245:85/eebhub