The Value of Data Modelling and Tagging for Smart ...€¦ · •Reference implementations: Java, node.js, Dart, Python, C++ others •Plug-ins to enable various systems to “speak”

a Project Haystack educational event Sponsored by

The Value of Data Modelling and Tagging for Smart Buildings

- an introduction to Project Haystack

Chris Irwin

- Project Haystack Executive Envoy for Europe

also

- VP Sales EMEA & Asia + VP Global Marketing for J2 Innovations


Our Goal – making data from diverse systems easy to work with

project-haystack.org

marketing.project-haystack.org


The Challenge of Data Management

• Most data produced by devices & equipment is poorly defined as to what the data means

• No standardized approach to adding semantic data (metadata) to device data

• Names can’t do the job

• Manual, labour-intensive processes are required to add semantic definitions to data before analysis, presentation and other value creation can begin

• This is a barrier to effective & efficient use of device data – adds cost and slows the use of data


Why Project Haystack – the data paradox

On the one hand, data seems to be everywhere, in every system ready for the “picking”

However, although everyone is asking for data, sometimes it is without understanding it, or how and where it should be applied or how it affects workflow and business operations

There’s the joke that goes….

“80% of data is about organising the data, and 20% is complaining about having to organise it.”


Unifying Data

• Data that matters

• Quality & integrity

• Tagging & Modelling

• Governance

• Portability

• Consistency

• Best practice

• Understanding

• Normalisation

• Socialisation

• Management

• Uniformity

• Interoperability

• Exchange

www.datafoundation.org

Factors to consider in a unified data standard – according to the Data Foundation


The Solution

• Haystack tagging unlocks the data from the BMS and other systems

• Standardised semantic data enables many applications to use the data - automatically


Project Haystack is…

• A community of people working to address the need for a standard “semantic modeling” method for device data

• Open source – no cost to use

• Worldwide community

• Proven and deployed in tens of thousands of facilities


Value Creation

Tagging :

• Creates new opportunities for value creation for owner/operators, facilities managers integrators, and OEMs

• Ensures multiple users are using same version of the organisation’s data

• Systems and data sources are linked and managed consistently to ensure that any data used by the organisation is consistent & accurate

• Safeguards that the right data gets to the right person at the right time


Project Haystack provides…

A standardised methodology for describing data that makes it easier and more cost effective to analyse, visualise, and derive value from our operational data

Think of it as a “MARK-UP LANGUAGE” for data


Analogy: A mark-up language for device data

• Why can I point my browser at your website and read what you published?

• We didn’t pre-arrange for me to be able to interpret your website code

• It only works because industry agreed on a mark-up language (HTML)

• If you use HTML I can read the “data” on your website (text)

• Haystack does the same thing for device data


The Project Haystack homepage the way we see it


The same page described in HTML


Defining data and its context

• Analyze this: zn3-wwfl4 = 21.2

• Hmmmm… What does the number represent? Deg C, F, KW, kPa???

• Need to know units. Lets say it is Deg C

• Hmmmm… is 21.2 DegC OK?

• What is it? Zone temp, Return air temp, Chilled water temp? Lets say it’s a Zone

• What is the schedule for the space? Schedule #1 = 7:30 AM - 6:30 PM

• What AHU is it served by? AHU-1

• What FCU serves it? FCU-27

• How can I convey these answers in a standard way that other software can interpret?


Standardising the metadata – 3 types

Defining data and its context


But Project Haystack is more than just about tagging…

• 1st – a standard methodology for defining and representing device metadata (descriptive data) – a specification on how to do it

• 2nd – Standard vocabulary (tag sets), taxonomies (equipment models) developed by consensus of the community


But Project Haystack is more than just tagging…

• 3rd - Software Tools• REST API to easily exchange Haystack tagged data

among applications

• Reference implementations: Java, node.js, Dart, Python, C++ others

• Plug-ins to enable various systems to “speak” Haystack

• Tools to streamline the tagging process

• 4th – Ongoing effort by Working Groups to develop tagging models, extend the standard, working with other standards groups, and educating the market


The Pay-off – what it enables

• Applications that just work!• Example: equipment graphics that auto-generate just by reading the metadata

associated with points

• Control logic can ”find” all similar devices it should be applied to (think of room controls or FCUs)

• Easier integration among software applications • apps can understand and consume data without human interaction to “map” data

• A new generation of engineering tools to streamline project implementation tasks


Haystack – adoption and support

• Used in systems and software deployed in tens of thousands of buildings to model hundreds of thousands of devices • last count – well over 30K facilities!

• Adoption by equipment manufacturers for next generation product – some on the market today

• Intel joins Project Haystack as a Board Member (March 2017)

• Many systems integrators trained and using Haystack in projects every day

• Over 1,600 registered users on Project Haystack Forum

• CABA White Paper March 2016


Project Haystack – our members


BACnet Collaboration with Haystack

ASHRAE Standard 223P: "Designation and Classification of Semantic Tags for Building Data” provides a dictionary of semantic tags for descriptive tagging of building data including building automation and control data along with associated systems


Evolution of data management


Project Haystack v4 – in beta

(at public review stage since March 2019)

https://project-haystack.dev/


Project Haystack v4 – What’s new?

•Naming

•Classification

•Relationships



•Naming

VFDsEnergy

Zones

AHUs

VAVs

UnitaryEquips

Chillers

Boilers

Tanks

Electrical Panels

Networks

Lighting

Libraries

Security


lib::IoT

Naming – Libraries needed to identify different meanings

hot

lib::food

hot



•Classification


Classification is hard


English is our enemy

hot-water = hot + water

site-meter = site + meter

hot-water = hot + water

site-meter = site + meter X siteMeter



•Naming

•Classification

•Relationships



•Naming

•Classification

•Relationships

= Vocabulary

= Taxonomy

= Ontology


Why does any of this matter?

• Data portability & abstraction

• Reduce commissioning errors

• Powerful data queries

• Intelligent rules, moving to AI

• Intuitive configuration tools & UI software


Haystack – key take-aways

• Deployed, working, proven in THOUSANDS of applications

• Open source, community-driven, ZERO cost to access documentation and use

• Extensible beyond community agreed equipment models – you can use Haystack methodology with your own tags/descriptors outside of standard group work on models


Haystack – key take-aways (continued)

• Lightweight – can be implemented in small devices, network level controllers, standard databases – also text files, & Excel worksheets

• Human readable and machine readable

• Accessible/understandable by real users – technicians and engineers that do systems integration

• The standard, libraries and tools continue to advance through the efforts of a worldwide open source community


Any questions?

Chris IrwinJ2 Innovations

We look forward to seeing you next year at ExCeL London

Documents

The Value of Data Modelling and Tagging for Smart ...€¦ · •Reference implementations: Java, node.js, Dart, Python, C++ others •Plug-ins to enable various systems to “speak”