21 May 2014 Motivation Evolution of Building Management Systems - - PowerPoint PPT Presentation

21 May 2014

Project Overview

Motivation – Evolution of Building Management Systems

T

• day’s commercial buildings
• Wealth of difgerent, but isolated automation,

management and information systems

• Integration of the difgerent systems

requires high engineering efgort

Future integrated smart buildings

• Integration of legacy and novel BAS with

minimal engineering efgort

• Convergence of BAS and ICT systems
• Uniform Information Model and

comprehensive Semantic Knowledge

• Value-added building services on top of

integrated BAS and ICT systems

Vision – Future Integrated Smart Buildings

Overcome traditional silo structures

• no interoperability between BA domains
• hindering comprehensive data analysis
• no creation of knowledge

Avoid proprietary engineering approach

• not fmexible: high engineering efgort required
• proprietary: lack of standardized interfaces
• no quick creation of new applications

Introduce common semantic model

• system to be operated by experts only
• no common information model and

semantics

• no easy integration of existing / legacy

systems

• no quick creation of new applications

BaaS Challenges

BaaS Approach – Situation and Vision

Situation: A building consists of difgerent systems with difgerent

• Responsibilities:

–

Building automation: heating & ventilation, lighting & blinds, …

–

Facility management: room reservation, access control, …

–

IT management: network access, printing services, beamer control, …

• Partly standardized data models: BACnet, KNX, oBix, OPC-UA, …
• Proprietary engineering models: Siemens DESIGO, Kieback & Peter, ...
• Communication interfaces: KNX, BACnet, EEBUS, CAN, WS + oBIX, ...
• Degree of self-description: the integration of BIM + BAS + BAS-semantics is currently not existing

Vision: Integrated building systems that are easily extensible in terms

• f new services and applications and are based on an open standardized semantic building

services platform.

BaaS Approach – Problem and Mission

Problem (what we sufger from)

• Engineering, commissioning and operation of domains requires a host of difgerent tools (from

difgerent vendors, for difgerent technologies)

• The resulting system models, confjgurations and interfaces are mostly proprietary and often not

available online or in the fjeld

• Lack of common information models and integrated engineering tools makes it very hard to add

additional services to an existing installation Mission (what we need to do)

• Standardized communication mechanisms (a.k.a. transport bindings)
• Domain independent data models and associated representations
• Semantic models for services and data following common ontologies
• T
• ols for integration of legacy systems, data models, and interfaces
• T
• ols for developing and integrating new value-added services and applications (incl. engineering

and commissioning support)

IoT Architecture Reference Model: Functional View

IoT ARM: Deployment Confjgurations

BaaS Reference Architecture (Functional sketch)

European Project Partners

21 May 2014

• Model based management
• Planning phase
• Run time phase
• Abstraction layer at run time
• Hierarchical object model for management of relevant resources

Project partners

9/1/14 Materna 11

• Model based management
• Usability
• Event-condition-action
• OSGI remote services
• Legacy support

Project partners

9/1/14 TU Dortmund 12

• Real time capable web services
• Web services on embedded devices
• Real time web service

Project partners

9/1/14 Universität Rostock 13

• Domain specifjc languages
• Platform SDK
• Semantic web of things

Project partners

9/1/14 Siemens 14

• Smart city
• Smart lighting
• Light paths for humans
• Room booking
• User feedback, e.g. NFC

Project partners

9/1/14 everis 15

• Value added services based on geographic information
• Geolocated objects and services
• Indoor positioning
• Visual interaction with the building

Project partners

9/1/14 prodevelop 16

Smart Meeting Room

21 May 2014

• Lights
• Heating, Ventilation, AC
• Shutters

and enables convergence with traditional IT:

• Network
• Computing devices

Smart meeting room

9/1/14 18

The Smart Meeting Room orchestrates beyond silos:

• Presence detection utilizing a variety of sensors
• Passive infrared
• Ultrasonic
• Identifjcation
• RFID
• External conditions
• Light
• T

emperature

• Room planning and schedule

Reconfjguration triggered by…

9/1/14 19

+

• T

enant profjles and policies

• User feedback
• User corrects mistake, e.g. adjusts temperature
• Optimization and predictive automation
• Modularized reasoning over context

Reconfjguration

9/1/14 20

21 May 2014

Predictive Automation

• Goal: develop mechanisms to reduce the energy consumption of

shared spaces, while still complying with the needs of users

• Example: proactive temperature control
• f a meeting room based on user-preferences, available devices, room

layout, outside temperature and historical data

• T

echnical Approach: sensor fusion & information processing & domain-specifjc optimization algorithms

• Main BaaS features needed:

Semantic abstraction from actual hardware

Use-Case: Predictive Automation for Energy Effjcient Buildings

BaaS – 21.05.2014 Andreas Müller – TWT GmbH 22

Semantic BaaS framework

Room Bookin g Syste m User preference s

Shared Space

Historica l data

TWT Optimization Algorithms

control