This document discusses standards and protocols that enable interoperability for connecting devices to the internet of things. It notes that while there are many choices for communication protocols, interoperability requires agreement on information exchange standards. The talk will cover existing information exchange standards and technologies that can be used with the standards. It emphasizes that every API or protocol standard should have an accompanying information model that defines relevant concepts to provide necessary context and metadata. Semantic web technologies like ontologies, RDF, and SPARQL can help integrate data from different sources by mapping terms and relationships between standards. This approach supports interoperability across distributed, loosely coupled smart city systems.

1. © 2013
Connecting the Next Billion Devices
to the Internet
Standards and Protocols
Steve Ray
Steve.ray@sv.cmu.edu
IoTX 2015

2. IBM Research
2
 You are hearing about the benefits of smart cities
and the Internet of Things
 One essential assumption: Interoperability
 This talk is about the plumbing that makes it
possible
– Information exchange Standards
–Technologies that can be used with these
standards
2

3. IBM Research
3
Connectivity ≠ Interoperability
Lots of choices of
Information content
IP Layer Strong agreement here
Lots of choices of
Communication Protocols
Zigbee
Wifi
Power line
Not so much attention
Lots of attention
3

4. IBM Research
4
What’s the problem?
 Misunderstanding of information flowing between systems
“Set Thruster to 324.59”
September, 1999
4

5. IBM Research
5
Context
 Many standards provide context through
supporting documentation for human
implementers to read – and possibly misinterpret.
 A good information standard explicitly captures
metadata (context)
– Information model, or
– Semantic model, or
– Formal logic model
UML
OWL
First order logic
5

6. IBM Research
6
Need context (a.k.a. metadata)
“Thruster setting to Impulse with value 324.59 Newton-seconds”
Thruster settingImpulse
Newton-second 324.59
6

7. IBM Research
7
 Every API or protocol standard should be supported
by an information model that defines all the relevant
concepts
 “Need smart data so you don’t need such smart
software”
– Krzysztof Janowicz , UCSB
7

8. IBM Research
8
Whenever you see something like this:
8

9. IBM Research
9
You should have access to something like this:
9

10. IBM Research
10
10
Part of ASHRAE
Standard SPC201
“Facility Smart Grid
Information Model”
(FSGIM)
class Device
Generator Component::GeneratorLoad Component::Load
Device
+ deviceType: MasterFormatType [0..1]
+ extendedInfo: ExtendedInfoType [0..1]
+ status: DeviceStatus [0..1]
ComponentElement
Meter Component::Meter
«enumerati...
DeviceStatus
Normal
Warning
Alarm
NAESB PAP10 EUI::
NameType
+ name: String
NAESB PAP10 EUI:
:
NameTypeAuthority
+ name: String
MasterFormatType
+ code: String
ExtendedInfoType
+ ExtendedInfoReference: String
FSGIMIdentifiedObject
+ aliasName: String [0]
+ mRID: GlobalId
+ name: String
+ nameType: String
+ nameTypeAuthority: String [0..1]
NAESB PAP10 EUI::
IdentifiedObject
+ aliasName: String [0..1]
+ mRID: String [0..1]
+ name: String [0..1]
NAESB PAP10 EUI::
Name
+ name: String
CDCDescription::DPL
+ vendor: String
+ hwRev: String [0..1]
+ swRev: String [0..1]
+ serNum: String [0..1]
+ model: String [0..1]
+ location: String [0..1]
+ name: String [0..1]
+ owner: String [0..1]
+ ePSName: String [0..1]
+ primeOper: String [0..1]
+ secondOper: String [0..1]
+ latitude: Real [0..1]
+ longitude: Real [0..1]
+ altitude: Real [0..1]
+ mrID: String [0..1]
Energy Manager Component::EM
+ hasLoads: Boolean
+ hasElectricalGenerators: Boolean
+Names 0..*
+IdentifiedObject 1
1
+loads
0..*
+deviceNameplate
0..1
+eMs 0..*
1
+meters
0..*
+Names 0..*
+NameType 1
1
+generators
0..*
+tags
0..*
+NameTypes 0..*
+NameTypeAuthority 0..1
+component
1..*
10

11. IBM Research
11
Information models for city systems
http://www.ibm.com/developerworks/industry/library/ind-smartercitydatamodel1/index.html
Standard Examples of supported concepts Current deployments
Common Alerting Protocol (CAP) Category, status, scope, certainty, severity, urgency,
onset time, expiration time, response type,
instructions
International standard (OASIS plus ITU-T
Recommendation X.1303) with deployments primarily
in the U.S., which include: Department of Homeland
Security, National Weather Service, United States
Geological Survey, California Office of Emergency
Services, Virginia Department of Transportation, and
Oregon RAINS.
National Information Exchange Model (NIEM) Activity, address, case, date and time, document,
item, incident, location, organization, person
U.S. specific, with deployments that include: U.S.
Department of Homeland Security, U.S. Department
of Justice, U.S. Citizenship and Immigration
Services, Federal Emergency Management Agency
(FEMA), Law Enforcement Information Sharing
Program, Logical Entity eXchange Specifications,
OneDOJ, and the Department of Health and Human
Services.
OpenGIS Geography Markup Language (GML) Point International standard (OGC) that is widely used in
the industry, and considered the reference in its
space.
OpenGIS Location Services (OpenLS) Location International standard (OGC)that is used for location
based applications such as cell phone apps.
SOA Ontology Service, process, task, event, human actor, effect,
system, policy, service contract, service interface,
element
International standard (TOG) that encapsulates SOA
vocabulary and relationships and is used to describe
and model SOA solutions.
Universal Core Entities and assets, events and alerts, people,
organizations, locations, collections
U.S. specific standard that is jointly managed by
Department of Defense, Justice, and Homeland
Security, and the Office of the Director of National
Intelligence. Within the DoD, the Marine Corps, Navy,
and Air Force appear to be committed to supporting
UC.
W3C Time Ontology in OWL Interval, DurationDescription, DateTimeDescription,
DayOfWeek
International standard (W3C) with and uncertain
adoption. Our web search revealed no concrete
deployments.
11

12. IBM Research
12
 Organization
– Definition: A group of persons organized for a particular purpose
– Examples: Police department, public housing department, bus department,
transportation agency, water agency, electric utility
– Key attributes: Name, type of organization, description, identification, website
– Key relationships: Organizations (parent-child), assets, location
– Standards assessment: National Information Exchange Model (NIEM): NIEM-
Core (nc:)OrganizationType, UCore (Universal Core) organization
 Alert
– Definition: A warning or alarm for an imminent event
– Examples: Road repair advisory
– Key attributes: Sender, description, urgency, severity, certainty, onset time,
location, supporting resources
– Key relationships: Sender (organization or person), location, incident, work
orders
– Standards assessment: The Common Alerting Protocol (CAP) has extensive
support for the alert concept. The UCore Event concepts are also applicable.
 Incident
– Definition: An occurrence or an event that may require a response
– Examples: Road repair, automobile accident, water main bursting, criminal
activity
– Key attributes: Date and time of the incident, description, ID
– Key relationships: Location, alerts, work orders, owner (organization or
person)
– Standards assessment: NIEM:nc:IncidentType, CAP:alert:incidents, UCore
Event, Service-Oriented Architecture (SOA) Ontology Event
 Person
– Definition: A human being, an individual
– Examples: James, Bob, Sally
– Key attributes: Full name, given name, family name, gender, date of birth,
place of birth, citizenship, country of birth
– Key relationships: Employer, location, address, organization, role (such as
operator, supervisor, responder, analyst, asset manager)
– Standards assessment: NIEM:nc:PersonType, UCore Person, SOA Ontology
Human Actor
 Asset
– Definition: A tangible object that can be tracked over time
– Examples: Road, water pipe, electric capacitor, bus, building
– Key attributes: Description, ID
– Key relationships: Organization, person, manufacturer, location, work order,
incident
– Standards assessment: NIEM:ip: AssetType, UCore Entity
 Work order
– Definition: An order to do some work; to fix, repair or replace
– Examples: Road repair, utility maintenance on a main valve, re-routing of buses
– Key attributes: Description, ID, comment, priority, status, location, start date/time, stop
date/time
– Key relationships: Work steps, work orders (parent-child), incident, alert, organization,
maintenance history, specification, person, assets
– Standards assessment: No relevant standards identified at this time.
 Process and procedure
– Definition: A series of actions to accomplish a goal
– Examples: Road repair notification and coordination
– Key attributes: Process document
– Key relationships: Process steps, work orders, incident, alert, organization, person, assets
– Standards assessment: SOA Ontology Process
 Key Performance Indicator
– Definition: A measurement or criteria to assay the condition or performance of a person,
process, or thing
– Examples: Response time, time to closure, cost to city, savings to city, impact to city
services
– Key attributes: Description, metrics, thresholds
– Key relationships: KPI (parent-child), organization, incident, alert, process and procedure,
asset
– Standards assessment: No relevant standards identified at this time.
 Location
– Definition: A geographic place, point, position, or area identified by its coordinates in an
earth based coordinate system, name, or address
– Examples: Road repair location: city intersection, water pipe location
– Key attributes: Geographic coordinates, postal address, TimeStamp
– Key relationships: Person, organization, asset, incident, alert
– Standards assessment: NIEM:nc:LocationType, UCore Location, Geography Markup
Language (GML) Point, OpenGIS® Open Location Service (OpenLS) Location
 Time
– Definition: Measuring system used to sequence events, to compare the durations of events
and the intervals between them, and to quantify rates of change such as the motions of
objects
– Examples: Start time, end time
– Key attributes: Years, months, weeks, days, hours, minutes, seconds, milliseconds
– Key relationships: Duration
– Standards assessment: NIEM:nc:DateTime, W3C DateTimeDescription
Source: Rick Robinson, Executive Architect, Smarter Cities, IBM
Some Relevant Concepts
12

13. IBM Research
13
Some Smart Grid Standards
CIM 61968
CIM 61970
CIM 62325
61850
ISO 16484
BACnet
61850-410
hydro
Multispeak
C12.19
Abstract Model – Shared Concepts, Fragments???
61850-420
DER
(solar…)61400-25-2
wind
61850-420
DER
(solar…)
WS Calendar
NAESB
energy usage
info
EMIX
OASIS energy
interop
Zigbee smart
energy profile
ISA88
ISA95
CEA 709
LonTalk
62351-7 comm net
and system mgnt
ICCP
IEEE 1815
dnp3
IEEE c37.239
comfede
naspi
ASHRAE
SPC201
FSGIM
OpenADR
13

14. IBM Research
14
IEC 61968
IEC 61970
NAESB PAP10
Multispeak V4.1
“Meter”
Do they all really mean the same thing?
14
14

15. IBM Research
15
What to do?
 Require all standards to use the same
universal definitions for all terms
 Require all standards to explicitly define all
terms (in machine-readable form)
 Use technology to automatically map
between terms in one standard to terms in
another



15

16. IBM Research
16
Semantic data integration for Internet of Things
CMUSV
Sensor Data
NASA
Sensor Data
NICT
(Japan)
Sensor Data
•Retrieve data
•Import or synthesize ontologies
•Map to general ontology
Technologies
RDF
OWL
SPARQL
SPIN
Also Enables
Inferencing over data
Federated data stores
Distributed queries
….
16

17. IBM Research
17
Base Ontology
17

18. IBM Research
18
Extension Ontologies for Specific Domains
18

19. IBM Research
19
Linked Open Data Cloud
19

20. IBM Research
20
20

21. IBM Research
21
Querying DBpedia
21
SELECT DISTINCT ?buildingName ?floors
where {
?building a ?o .
?o rdfs:subClassOf ?class.
FILTER (regex(?o, "dubai","i")) .
?building dbpprop:floorCount ?floors .
FILTER (?floors > 80) .
FILTER ((datatype(?floors)) = xsd:integer) .
?building dbpprop:name ?buildingName .
}
ORDER BY ?floors
(Show all entries for buildings with more than 80 floors)

22. IBM Research
22
Interoperability Standards battles
AT&T, Cisco, IBM, Intel…
plus ~165 others
Intel, Samsung, GE, Cisco, Dell…
plus ~50 others
Microsoft, Cisco,
Qualcomm, Sony, LG…
plus ~140 others
Thread Group
Google (Nest), ARM,
Samsung appliance…
plus ~120 others
Apple Homekit
“dozens of partners”
as of May 14, 2015Hewlett-Packard, Cisco,
IBM, Oracle, Philips
Plus ~400 others
22

23. IBM Research
23
Takeaway for Smart Cities
 We need to share information
across different contexts
 Information needs to be defined in a
way that machines can understand
the meaning
 A neutral model and semantic
mappings give us a way to manage
complexity
 Semantic web technology is well-
suited to loosely-coupled,
distributed, linked communities of
systems
23