See also: IRC log
<kaz> scribenick: kaz
oliver: CEO of lemonbeat
... happy to host this event
remote+ Zoltan_Kis(Intel)
oliver: [The Lemonbeat GMBH]
... origin is smart home
... happy to contribute to IoT space
... would introduce Andreas
UNKNOWN_SPEAKER: [BMWi* Funding
programme]
... 2 step approach: sketching and actual project
... [The 5 "SINTEG" showcases]
... [know-how for the region]
... we can cover all the energy transitions
... high-load areas, medium-load areas and low-load areas
... combining 3 different areas
... getting funding
... [Experienced partners from on of the economically strongest
regions in Germany]
... 31 project partners, 15 associated partners
... [The energy system becomes more distributed and complex]
... in the past - in the future
... [The distribution network - Backbone of the energy
transition]
... EHV - HV - MV - LV
... more than 90% of the renewable energy systems are fed into the
German distibution network
... [Sokutions which ensure that the region actually benefits from
the energy transition]
... huge infrastructure
... self-learning by AI
... out of experience
... [Designets integrates individual solutions from various
sectors]
... overall system, e-mobility, intelligent grid, digitization,
distributed generation, flexible industry
... no standard protocol yet
... standardization would be very helpful
... [Just ike DNA in cells, our lemonbeat software stack with our
LsDL empowers each device]
... offer flexibility and get flexibility from the market
... [We make the energy transition easy to grasp with the "ENERGY
ROUTE"
... demonstrators come alive
... signage of demonstrators along the "energy route"
... access to demonstrators, "showrooms", events and open-house
days
... combination of different projects
... sector coupling for different energy transitions
... [Five answers]
... our motivation, our goal, our approach, our path, our
showcase
... need to be neutral to various vendors
... tx!
... questions?
lindsay: web protocol working so far?
and: lemonbeat implementations?
scribe: yes, there are
dsr: electric car is one challenge, and...
and: e-mobility massively effects our infrastructure
scribe: shift the systems
... 2 opportunities
... challenge is renewing the end
... huge tech problems
... e.g., voltage drops
... infrastructure like cables
... using IoT compensate usual consumpation
mk: how many different protocols/standards are there?
and: can share my feeling :)
scribe: the solution provides
flexibility
... heat pump, heating system, water system, etc.
... not based on standards so far
... there are different manufacturers
... using different protocols
... standard interface would help interaction
... if you want please follow the slides and contact me
... tx!
uday: meeting followed by a social
event in the evening
... city tour and dinner
... looking forward good discussions
<dsr> scribenick: dsr
Kaz Ashimura (W3C) introduces the W3C Patent Policy and its implications for our work.
W3C advances specifictions through 5 steps from an initial Working Draft to the final Recommendation (W3C’s term for its standards)
W3C Recommendations must be available to implement on a royalty free basis.
You are encouraged to join the Working Group if you haven’t done so already, and to make the corresponding licensing commitments.
The WG Chairs can invite guests to WG meetings, however, we need to be aware about potential IPR issues that could arise from contributions from people who are not formally part of the WG.
Kaz shows the important points from the W3C Patent Policy document
When contributions are being considered for inclusion in a document intended to become a W3C Recommendation, the WG Chairs should ask the contributor to disclose any essential claims that they would not be able to provide royalty free licenses as per the W3C Patent Policy
Kaz lists the managed repositories relevant to the WoiT WG
If the WG is unable to secure licensing for essential claims under the W3C Patent Policy, the Chairs should endeavor to steer around the concerned technical aspects.
Michael Koster (SmartThings/Samsung) presents.
There are many standards organisations, and these have taken their own approach, and in some cases are competing with each other.
The approach we’re taking is to enable semantic interoperability at the application layer
We’re looking at well known formats to describe how applications can interact with things.
We’re adapting the patterns from schema.org that were developed for smart search results, and combining this with other ontologies and vocabularies.
We want to do this in an open process that doesn’t require formal membership of a standards development organisation..
Contributions based upon either the W3C Community Group or the schema.org frameworks
Michael presents a semantic stack model
This has the ontologies at the top, the web of things in the middle, and specific IoT standards suites at the bottom
We want to enable a means to abstract capabilities, e.g. motion sensing, turning something on and off and so forth.
Actuation may be incremental, e.g. light levels and colours rather than simply on and off
Michael describes how the iotschema,org capability pattern relates to the web of things.
Essentially a capability is mapped to an interaction pattern with its associated properties, actions and events.
Michael shows a JSON based representation of an example capability
We’ve done some initial work to prepare a charter. We hold monthly teleconferences and we’re about to apply the charter to the W3C Web of Things Community Group.
We plan to use the CC-SA license for contributions.
Collaboration with the W3C WoT IG and WG
We’re aiming to collaborate with other organisations to enable re-use of existing ontologies
Work on factoring out differences across ontologies, semantic annotations and support for composing services.
Zoltan: I really like the notion of capabilities
Michael: At Smart things, we have had several years experience with capabilities
Capabilities also help with security
<zkis> capabilities could be used as funcionality units for permissions and policies
Capabillities are building blocks for things and not for use on their own
We don’t need to express capabilities directly in thing descriptions, but rather to use them to describe relationships between things and capabilities
Stefan: is this approach already accepted by schema.org ?
Michael: that’s an interesting question. We’re really at an early stage where we are working on a proposal that we want schema.org to take on board
Matthias: in the Web of things, we want to model the physical world. We need to relate the interaction model to the physical capabilities
Dave: we need to recognise that different communities will develop different vocabularies, so we need bridging ontologies that allow search using shared concepts rather than having to use the different terms for each of the different vocabularies
Lindsay: ETSI is also working in this area. The problem of relating different models can be likened to the parts of an iceberg that is under the water - it is very much larger and risks sinking the ship of interoperability
Philipp Hoschka: what is the relationship betweeen schema.org and W3C?
Tobias: how do you exploit these descriptions in applications?
Michael: the combinations of these different kinds of descriptions can be used to support discovery, composition and adaptation by smart services
The Web works because it is simple, much simpler than some of the older ideas for hypermedia
We want to make it easier for people to innovate and contribute new ideas
The protocol binding for the web of things maps the abstract models to the concrete protocols
We want to avoid the complexity of the Semantic Web and make life easier for regular developers.
Michael_McCool: there is a risk of alienating developers if we don’t do a good job on that.
Kajimoto-san: one challenge is what things call themselves, another is who owns the things.
Michael: we need a means to describe the interaction model and also a means to describe the relationships between things at a semantic level
There is a lot of other categories we need to look at including security policies
Luca: we should exploit the work that has been done for the Semantic Web, why aren’t we using RDF rather than JSON?
Michael: we trying to use the framework of RDF but to avoid the need for developers to first become semantic web experts
We have some choices to explore for this
Dave: we need to keep distinct the serialisations from the underlying framework.
<kaz> [break]
Presented by Juan Carlos Zuniga (Sigfox)
Juan introduces Sigfox. This is a company that focuses on low power wide area connectivity for IoT devices.
Key goals include low cost and low data rates
Sigfox is a connectivity provider that builds and operates the network with regional SNO operators
We charge for connecting devices to the Internet
Sigfox can be applied to a wide variety of domains, including assisted living, home alarms, smoke detectors, precision agriculture, tracking moving assets.
5 main business benefits.
Sigfox coverage in the USA, South America, Europe, South Africa, Japan and Australia
Sigfox is designed to scale to very large numbers of devices,
The sigfox foundation supports applications such as tracking protected species like rhino’s
Sigfox uses ultra narrow band and frequency hopping.
This relies on software defined radio technology, also know as cognitive radio
Messages are sent with multiple copies in different frequencies
Uplink messages have 12 byte payload, downlink 8 bytes
Up to 6 messages uplink per hour and 4 messages downlink
After sending, the device listens for a reply in a short window of time
Devices spend most of their time asleep with negligible power drain, allowing for years of battery operation.
Devices are pre-provisioned with their credentials. Standard APIs are provided for accessing data on common cloud platforms.
Keen on standards based interoperability.
Collaborative approach to security, and design for privacy
SCHC - static context header compression for sigfox and LoRa from LPWAN WG
Demos of IPv6/CoAP over LPWAN
Interoperability at the application layer - demos for IETF
Privacy focused on explicit personally identifiable information
Example: if a trash bin is used, you know that someone is in the house.
Proactive privacy by design principles.
Resilience to botnets and tempered with devices - blocking and blacklisting capabilities in the cloud
Different levels of security according to the criticality and type of application.
Michael_Koster: what do you mean by device autonomy?
Juan: devices can talk whenever they want, there is no need to synchronise clocks etc.
some discussion around IPv6 via header compression
<scribe> On-going work on layering CoAP on top of this.
We’re looking at supporting message fragmentation and reassembly
Matthias provides some additional context
Juan: our cloud APIs provide metadata that comes from the cloud for the device, but is not held on the device
IETF work as a collaboration across several technology groups, e.g. sigfox, LoRa and 2 others
Darko: what format are you using for metadata and how is it managed?
Juan: right now every device maker does their own thing. We are seeking ways to converge. We’re not pushing on a single approach, but want to encourge convergence
Slides(tbd)
Presented by Stefan Schmid (Bosch).
The project started 18 months ago and is half way through
The project name is an acronym for bridging the interoperability gap of the Internet of Things
See: http://big-iot.eu/project/
Photo of busy city street with cars, trams and pedestrians, and many kind of IoT devices
Each of the IoT systems form independent silos
The objective is to enable interoperability across these silos
We want to foster and incentivize collaboration via a marketplace of services
The marketplace includes support for monetizing services.
The marketplace is aimed at developers not end users
Example for a smart parking service where street sensors report which spaces are not in use
This could involve a mix of technologies and communication standards
We need to support evolution of services and technologies. How can we enable applications to take advantage of new capabilities?
Service providers and consumers with authentication, registration, discovery, subscription and accounting
The service offering description specifies the access control policy
We use JSON-LD for the descriptions, based upon the work by the W3C Web of Things IG
Stefan talks us through the different kinds of information in the description.
We provide a simple API that abstracts away from the details of the underlying platform.
We can provide a gateway service for accessing non Big-IoT platforms.
Likewise, we support proxies when it isn’t appropriate for apps to directly access devices.
Big-IoT is working witg Eclipse IoT for our open source strategy
We plan to contribute semantic models to schema.org
We expect to very soon have approval from Eclipse for our proposed open source project.
Demo of the smart parking service in Barcelona
Stefan talks us through how a company could add its parking lot to the service.
You can query for parking space providers within a given radius of a given location.
A new provider needs to register its offering with the marketplace
This involves customizing some Java code …
The end user app shows available places and pays for the parking charge.
Juan: do you see a single market or multiple marketplaces?
Stefan: many marketplaces, first because of the amount of data, and secondly due to regional and domain specific variations of requirements
Kaz: are you interested in a standard API for payments from web browsers for web based applications?
Stefan: this is a research project with limited resources. There are many options for commercialising our work.
Philipp_Hoschka talks about the Compose EU project which also worked on marketplaces for IoT services.
Stefan notes there are already many IoT marketplaces.
Matthias: how did you survey the developers for the marketplace requirements?
Stefan: we learned that different developers prefer different programming languages, and that Java isn’t a universal solution.
<kaz> [ break ]
<kaz> scribenick: kaz
Lindsay: charing ETSI ISC CIM
... experience is very diverse
... [Audience Calibration: Hands up who knows of ...]
... SF book of "1984" and "Future Shock"
... also "Shockwave Rider"
... and then "General Data Protection Regulation" by EU
Parliament
... starting in May 2018
... smart city/parking would require much private data
... companies are not good at this kind of regulations
... organization in breach of GDPR can be fined up to 4% of annual
global revenue or 20M Euro
... [Audience Calibration: Hands up who knows of ...]
... manipulating genes
... changing DNA structure
... whole bunch of scratch pads
... like RFID embedded in cells
... [cell genome "recording of event histories"]
... transilation for "1984" and wiki-leaks readers
... cell genome could be made to inpact "your" cells...
... [Outline of this talk]
... ETSI ISG work
... main work is how to make exchange semantic data
... how to interpret it
... what would happen government set requirements for smart city
trials?
... [Digital Transformation for Wellington, New Zealand]
... half as NEC and half as ETSI
... measuring the quality of city
... how to improve comfort order
... sensing overcloud, etc.
... using video data, etc.
... tracking people
... [Digital Transformation for Christchurch, New Zealand]
... traffic jams
... tracking number of people
... [Smart Cities ... and standards]
... [Standards help all stakeholders]
... Standards help
... voluntary/dynamic
... creates an ecosystem of experts
... reduce risks
... governments/citizens need to ference standards
... issues with execution risks, cost explosions, obsolescence,
privacy, physical risks and liability
... have to worry how to record provenance
... need to know about your data
... license, quality, freshness, etc.
... [IoT-related SDOs & Fora: "Formal" overview]
... De Jure: ITU-T, ITU-R, ISO/IEC ETSI, ...
De Facto: W3C, oneM2M, ...
McCool: OCF as well
Lindsay: [Smart City related Groups
in ETSI]
... ISG CIM (Context Information Management)
... oneM2M here in the middle of the picture
... IoT systems are annotated
... problem from me is having a lot of stakeholders
... [Currently >300 EU IoT/SmartCity RnD projects]
... examples of A-B
... all those working on IoT and APIs
... [Work Item Scopes of ETSI ISG CIM]
... 6 different work items
... https://portal.etsi.org/tb.aspx?tbid=i54&SubTB=854
... ontology to describe information model
... no. 7 is security and privacy
... [Privacy Issues and Exploring Solution ...]
... [ETSI ISG CIM: Problem space and solution space]
... mind map
... goo.gl/1zFPRz
... [Context Information Management: Issues]
... enlarged mind map
... what kind of identifiers to use?
... privacy by design is a hot topic
... need to have some usage data
... what if including privacy information?
... [Context Information Management: Privacy]
... you can find these by visiting goo.gl/1zFPRz
... using blockchain to secure the data
... serialization of the data
... possible attack, e.g., by undermining
... de sur SDOs here
... CEN-CLS/JWG-8
... ISO/IEC 27550
... OASIS PMRMv1.0
... ISO/IEC 20547-4
... etc.
... [Links]
... [Help on ETSI ISG CIM Discussion Issues]
... 1. requirements for exchange of cross-cutting context
information, to be fit for purpose in Smart Cities, S,art
Agriculture and Smart Industry use cases
... 2. How to ensure ISG CIM API is flexible to cope with all
desirable types of context data and instance data
... 3. Requirements for ISG CIM API to express the important kinds
of queries for heterogeneous databases of context information for
cross-cutting domains
... 4. Managing the identifier(s) for entities and referencing them
in API queries, even across legacy databases
... Ensuring privacy protection (fine-grained access control,
encryption) on specific context data, efficiently!
... different identifiers based on various stakeholders
... most of the systems are walled garden
... usually not available by external parties
... [How to help]
... Ask an expert on any of the 5 Discussion Issues to contact
me
... Download the CIM-mindmap goo.gl/1zFPRz and suggest to me your
additions
... Email me about other events/organizations which I should
contact
... Recommend to your organization to join ETSI iSG CIM
... thank you!
Koster: question@@@
Slides(tbd)
Benjamin: [Goals]
... Interest of WoT
... integrate data from the vehicle with external knowledge
bases
... adds a semantic enrichment layer
... can be linke with the work of the W3C AUtomotive WG (and
GENIVI)
... [General architecture and parts]
... WoT Servient on the left side
... OEM cloud on the right side
... HTTP between them
... and CoAP between cloud and car
... [Protocol binding: LwM2M]
... WoT methods -(Python snippet)- HTTP methods -(Leshan API)-
LwM2M methods
... [Demo setup]
... OBD dongle
... browser as a WoT client
... OEM cloud
... connected using HTTP
... protocol binding and finding the device
... [Thing description: context]
... need: sensor/actuator ontology, car ontology
... w3c-wot-td-context.jsonld
... w3c-wot-common-context.jsonld
... schema.org
... linked version of GENIVI's VSS
... [TD retrieval and parsing]
... the TD file is stored in the dame repo as the WoT server
... [Thing description: sequence diagram]
... changed from yesterday
... discovery -> consume TD -> read speed
... communication among browser, WoT server and LwM2M API
... [WoT communication]
... WoT server: python script with Flask
... WoT client
... [Final interface]
... API
... endpoint/consumed
... [Conclusion and future work]
... challenges faced
... remote access to vehicle data
... protocol binding between WoT and LwM2M
... link between W3C WoT and W3C Automotive
... future work
... questions?
McCool: translation to TD?
... automatically?
Benjamin: hand coded
kaz: using VISS or VIAS from the W3C Automotive?
Benjamin: VSS as the data model
Lindsay: subscribe the data?
Benjamin: can subscribe or get data at once as well
[ Everybody moves to the parking lot to see the demo]
[ PlugFest Day ends and everybody leaves for the social event ]