the onem2m base ontology - · pdf file•like sdt the onem2m base ontology provides a...
TRANSCRIPT
The oneM2M Base Ontology …and how this matters to HGI
Source: Joerg Swetina (NEC) vicechair of oneM2M REQuirements group
SDT from HGI
• HGI’s SDT (Smart Device Template) aims at a common approach for device modelling to enable interworking.
• SDT provides
– A set of automation commands,
– A common syntax,
sufficient to model most home appliance functions
https://github.com/Homegateway/SmartDeviceTemplate/
oneM2M Base Ontology
• oneM2M – like a HGI Smart Home Gateway – serves as an integration system that
– Enables interworking with external technologies
– Adds value (e.g. security, device management …)
• Instead of plain XML oneM2M uses ontologies (OWL) to describe the services, interfaces and functionalities from external technologies
– An ontology allows to: • Capture a shared understanding of a domain of interest.
• Provide a formal and machine manipulable model of the domain.
Why “ontology” ?
• Like SDT the oneM2M Base Ontology provides a template with very few essential concepts:
– A Thing: anything that can be described and identified in oneM2M
– A Device: special kind of Thing that can electronically communicate with other [oneM2M] entities. A non-oneM2M Device requires an Interworking Proxy for protocol translation
• has Services (and their functionality), DataPoints, Operations, Input / Output parameters
Some Base Ontology concepts
• Operation vs. Command
• Aspects, MetaData
• Things, ThingProperties, ThingRelations
A specific ontology is derived from the Base Ontology • Example: a standard KNX dimmer
– Functional Block – Dimming Actuator Basic (has been standardized). It contains KNX-DPTs:
• Switch, DPT 1.001
• Relative dimming, DPT 3.007
• Absolute dimming, DPT 5.001
Example: mapping KNX ontology into base ontology
“is-a” relationships of the “KNX ontology”
• A standard KNX dimmer is-a InterworkedDevice
• Functional Block “Dimming actuator basic” is-a Service
– “dimming-control” is-a Functionality (of that Service)
refers to an Aspect: “Electricity-Current”
• Operations – DPT 1.001
– DPT 3.007
– DPT 5.001
• Commands – Switch
– Relative Dimming
– Absolute Dimming MAS-2015-0618
Why “ontology” (…continued)
• Isn’t the base ontology just another kind of SDT? => Yes, but an ontology can do more:
• An ontology can be composed of multiple part- ontologies, e.g.
1. An ontology describing Devices, e.g. for home environment (sensors, actuators, controllers..)
2. An ontology describing Things, e.g. in a building (rooms, water pipes, electric outlets…)
• Together they can describe a complete system, including non-functional aspects (street, color..)
Why “ontology” (…)
• On oneM2M raw data can be ‘annotated’ – e.g. in RDF triple form – using a Semantic Descriptor that is based on an ontology.
– The Semantic Descriptor uses the vocabulary of the ontology to describe:
• the entity (e.g. a street in a smart city)
• entity relationship with other entities (e.g. connects to)
• The entirety of Semantic Descriptors can be searched, based on their meaning (SPARQL)
– E.g. find all streets with high CO2 concentration
Why “ontology” (…)
• Big Data processing, Analytics
– Highly improved by semantic description of data
• M2M service provider in the role of:
– Data broker
– Data re-seller
• E.g. anonymized
– Analytics provider
• Analytics in the edge
– System integrator
• “semantic connector”
Some challenges
• The world of semantics and the world of M2M are still at different levels
– Simple devices still provide just raw data, no semantic information
– In oneM2M the often data value needs to be fetched from a different place than its description.
– Semantic information is distributed. Discovery of semantics together with data values is complex, e.g.
• Find all sensors in a region XYZ that currently measure CO2-concentration with a value > 95
My expectations • HGI’s SDT initiative nicely harmonizes with the
oneM2M concept. It allows a simple way of providing technical information on devices.
• We need an ecosystem where manufacturers, experts and users are able to use a common terminology (=ontology) in a simple way.
• IoT will create “another type of Internet”, but this is a slow (not dramatic) process and requires much care by responsible people.