W3C

Automotive Working Group Teleconference

02 Jul 2019

Attendees

Present
Gunnar, Paul, Ted, Glenn, Magnus, Ulf, Daniel, Harjot, Benjamin, Don
Regrets
Chair
Ted
Scribe
Ted

Contents


F2F venue

Ted: we'll defer to email on this but I was curious if Patrick received an answer for hosting our F2F at VW in California. He was there a couple weeks ago and was going to try to get a firm answer in person
... if not we'll have to look at other alternatives, ideally from one of the participants in the group who have an office in the area
... I imagine people will want to book their travel soon

Workshop registration

Ted: second half of the week of our next F2F is the transportation data workshop
... in some cases this may be more pertinent to colleagues in your company which I hope you have reached out to to see if they are interested in attending
... even if far afield from your current work I would think it would be worth attending and you can report back to your company
... I sent a reminder to register to our mailing list earlier, deadline to do so is 8 July

Glenn: I will register and be submitting a whitepaper along the lines of the convergence accelerator

Daniel: same

Chapter 4 / data model

Ted: I was hoping for a quorum to complete this topic, there was maybe more of one last week in my absence

Ulf: last week the group was firm in their support on Daniel's wording in github issue

Daniel: I made progress since and would like to get feedback and issue a pull request

Ted: in the scattered conversations I think there are two main points, figuring out what belongs in VSS spec to make it more readable and standalone and then the wording for data model within Gen2 spec

Daniel: last week we had a few discussions on the nuances of the wording and feel we are pretty close. a concrete example from media or other domain would help

Gunnar: (via WebEx chat) Yes last meeting we discussed and achieved rough consensus on the infographic. Patrick, Laurent, Daniel, Ulf and myself. Ted, we even joked that you could be absent more often. ;-)

Ted: I should miss more calls due to travel :)

Ted believes second domain for data model under consideration from F2F was notifications with Laurent or Patrick taking a first pass

Roadmap

Ted: we didn't get to this at the F2F and thought we could potentially at least identify a couple next priority items while we work on fuller roadmap
...we'll wait for Patrick and defer

BG exmple application

Ted: Armin is getting closer to completion with his policy language and at the F2F wanted a clear, detailed example to see how to represent it
... we also had a demo of Ulf's Gen2 prototype with a speedometer client app
... in the BG data task force we discussed defining a mock 3rd party application that would want to access and evaluate a modest number of signals, perhaps off-boarding a few and then we can work on the accompanying policy
... we batted around a couple ideas eg vehicle management for a family, Glenn and Harjot were willing to talk to their product people to see if any ideas for an interesting/plausible one has come in from user feedback
... we'll put a writeup up somewhere and allow for input and scope creep from the group before freezing it

Harjot: I did reach out to a few of our contacts and waiting to gather the full reponse and will update you accordingly

Ted: it would be helpful to have some realistic if not real sample data but probably have to wait until we know what signals we will want to collect

Upcoming topics

several potential topics for upcoming calls such as BMW WoT demo, maybe better during our earlier call time in our biweekly rotation

Daniel: actually Benjamin and I were discussing earlier and could go through the slides and give an example now

BMW WoT demo

Minutes from W3C Web of Things workshop on demo

Slides

Paper

Benjamin: we attended the W3C WoT workshop hosted by Siemens
... the Things Description is advancing on REC track, it is a semantically annotated thing
... we chose HVAC system but you could pick any branch within VSS and we wanted to show some practical examples
... this slide is a serialization of the data. JSON-LD combines simple JSON information with linked data characteristers with corresponding ontologies you can dereference
... this makes it nice for machine and human readability. we did one with all the security requirements defined in the specification
... how you interact with the thing (vehicle in our case) what the payload of a message being sent to the car
... what protocol is being used. we built it on top of BMW connected drive, used the web API as a basis to build the car web thing servient (WoT term)
... this provided an interface showing flows in NodeRED of certain interactions
... they are connected to either web things end points
... we could include some nodes that are indirect interactions with the vehicle

Daniel: an important addition is that from the thing description there is no additional work needed to generate interaction rules in NodeRED (open source tool)

Benjamin: that is a rather complex example, when you perform this action, trigger this other...
... we showed this to the WoT community, we also played a bit with some more complex thing descriptions
... we received some good feedback
... non-automotive experts could devise ideas on how to utilize a vehicle as a thing

Daniel: (projecting making car things on fly) this is a view in nodeRED, car imported through thing description and here are some example flows
... we could read position long/lat separately, whether the car was secured (locked) and could perform remote unlock instructions
... we could use MS flow to send a message to your phone if an event occured on your vehicle
... the server nor car are running right now so cannot show that right now

Ted: I presume this is purely research and no immediate plans to turn into a product

Daniel: exactly, see how to bring something complex like a car to WoT

Benjamin: we basically used some more advanced features of thing desciption to link different parts (VSS branches) of the vehicle, how they relate

Gunnar: (via WebEx chat) No questions from me on the web of things, interesting though.

Ted: very cool, from earlier discussions I was expecting a vehicle to potentially be split apart into various simpler things for different purposes
... regarding longitude and latitude from elsewhere - maybe we want to add to VSS since we don't have a LBS/Nav domain underway
[better would be to come up with a rudimentary domain for long/lat etc and that can be discussed at transportation workshop]

Benjamin: it might not just be long/lat that is missing from VSS

Glenn: I was curious for autolock based on distance from vehicle, what is the communication like?

Benjamin: I didn't work on that piece of the code, think it was just a distance calculated from the app on phone

Daniel: correct, we sent location from phone to nodeRED server that knew where the vehicle was and did the calculation
... we didn't worry about battery/power consumption or anything

Glenn: instead of VISS on the car you are getting the data from BMW connect and converting?

Benjamin: correct, you could produce the signals otherwise and map to VSS and use in nodeRED
... to Ted's earlier comment, yes an expert to a subset of signals and concerns around them would think of solutions and consumer needs
... we can customize the behavior and preferences of the customer as their needs may vary
... as you said, it isn't about exposing a vehicle as a thing in entirety but one of several things for different purposes

Ted: cool, so what's next? - that you can talk about obviously

Daniel: this demonstrates the power you can achieve with these standards and using nodeRED shows that you can rapidly drag and drop custom capability
... next step is transportation workshop
... this gives us again an opportunity to see how it can fit into broader transportation realm
... the domain knowledge we express in VSS/VSSo can be combined with other transportation ontologies for a wide array of purposes

Glenn: it has got me thinking about how we can integrate other applications and create value
... this may also help draw other OEM into these standardization conversations

Daniel: disclaimer: just because we express interest in standards doesn't mean we are opening this up anytime soon
... it is still just a research project and there will be business decisions as for what is next. we are thinking of the technical perspective

Glenn: understood. looking from OEM perspective, open would be risky and a gateway to services would be necessary to govern interactions

Ted: from what I understood NodeRED was operating in the cloud and communication go-between from car and other thing (smartphone etc), no direct interaction to vehicle
... strong auth can be enforced and conditions from vehicle (eg it is moving) that limits or severs tie to server

Daniel: correct

Summer Schedule

Ted: I'll start a doodle and send mail about summer schedule, figure out when people will be on vacation and we won't meet

Summary of Action Items

Summary of Resolutions

[End of minutes]

Minutes manually created (not a transcript), formatted by David Booth's scribe.perl version 1.154 (CVS log)
$Date: 2019/07/03 13:52:03 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
This is scribe.perl Revision: 1.154  of Date: 2018/09/25 16:35:56  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: Irssi_ISO8601_Log_Text_Format (score 1.00)

Present: Gunnar Paul Ted Glenn Magnus Ulf Daniel Harjot Benjamin Don
No ScribeNick specified.  Guessing ScribeNick: ted
Inferring Scribes: ted

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 02 Jul 2019
People with action items: 

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.


WARNING: IRC log location not specified!  (You can ignore this 
warning if you do not want the generated minutes to contain 
a link to the original IRC log.)


[End of scribe.perl diagnostic output]