W3C

Automotive data task force

04 Oct 2018

Attendees

Present
Armin, Benjamin, Glenn, Harjot, Ted, Hira
Regrets
Chair
Glenn
Scribe
Ted

Contents


TPAC

TPAC planning

Ted seeking ideas on individuals from WoT to try to meet with

Benjamin: Dave Raggett, and Michael Coster who has done some work with OCF on automotive

Ted: we should see if we can get some input from W3C Privacy Interest Group, certainly their Permissions UI workshop seems relevant to capturing consent

Ted to get details from Sanjeev and Wonsuk on their auto OCF PoC

Ted to get Privacy UI workshop notes

Main topic for F2F: (feel free to add to agenda wiki

* Consent capture

* Policy language, better understanding of LPL, alternates

* Privacy UI

* WoT / JSON-LD

* Data usage / marketplace

Benjamin: schema.org people not going to be there?

Ted: some
... not sure who, don't see eg Danbri

Benjamin: the auto schema.org people perhaps

Ted: I'll look at the participants and compare TPAC registrants
... and will see if Eric Prud'hommeaux will be in attendance

Data contract

Ted: from past discussion Data Contract was mostly focused on the metadata about how the sampling was done but there are more elements you can elaborate on

Glenn: issue that has arisen and becoming larger is global commercial fleets that require a telematics platform as part of their core business
... it is used for all sorts of purposes to manage fleet, including payroll
... there are very detailed specifications when purchasing large number of vehicles, tire size, engine type etc but not their data needs
... some fleets have been ordering thousands or tens of thousands of newer, replacement vehicles and finding there is less of the data they rely on exposed
... they have based their expectations and business model on previous vehicle capabilities
... it is a divergent group, global fleet base
... it ranges from class A trucks to light vehicles and encompass a growing number of vehicles
... it is not an easy thing to create a specification for the data. we are communicating with these fleet companies, courier trucks, rental companies, etc, and coming up with a template for a data spec
... we are of the opinion that we should be technology agnostic but their core needs must be met
... we hope to have a collective statement from this audience of large fleet companies from GAVDA, hopefully before TPAC

Ted: for heavy vehicles they also have a data port (Glenn clarifies different in EU and NA) and consistent data signals, not varied by manufacturer
... restricting access makes sense for OBD2 since allowing writes to CAN on running vehicles is a liability. some are starting to restrict the port but should generally continue to allow read only access during this transition period

Glenn: you are correct, there is a legitimate concern there and active discussion with manufacturers
... this transition period is causing problems for fleet owners and service providers

Ted: can you elaborate or provide examples on the signals that use to be present?

Glenn: seatbelt signal is important for commercial fleets, they want to ensure driver safety as there is significant financial impact from injuries
... that is no longer available for one manufacturer
... OBD2 is regulated by the EPA in US/North America and similar regulations in EU
... data that is suppose to be made available through that port is being restricted now
... we have been making manufacturers aware of these potential violations
... in these changes we need to be cautious not to remove data that is expected, agree on read only nature as well

Ted: OBD2 was originally intended for checking emissions data on running vehicles as you described. it became a convenient port for mechanics to check and clear diagnostics codes but otherwise should not have permitted writes

Glenn: that was the initial design and a marketplace came into existance based on the capabilities
... during this transition we need to ensure data access persists for these businesses
... in one case a manufacturer removed the data and provided an alternate. they do sampling every 1/2 minute
... they only do regular sampling at fixed intervals and miss any changes which might be important
... this is excessive, expensive and does not necessarily meet the needs of the data consumer
... commercial fleets now expect data to be provided differently
... the proposed equivalent is not that

Ted: difficult for individual fleet managers to communicate to OEM, good if their needs are conveyed. what sort of numbers?

Glenn: many millions, BMW estimates 1/6 of their vehicles alone
... there is a transformation occuring in how consumer receive goods and that is creating an increase in light trucks for last leg of shipment
... they are relying extensively on telematics data
... want to have clear means and ways, in standards, agreed upon

Ted: I am of the strong opinion that any attempt to create a massive silo of data that would meet all potential needs is impossible. I understand from the company that provides insurance dongles in NA that the data points of interest to their actuaries differ by company. this is within the same industry and same general purpose. clear you need to permit sampling on-board for bits pertinent to your needs and that is

Glenn: some data you only need for a second or two, such as when starting a vehicle a voltage drop could indicate battery or starter may be in need of attention. a time based sampling would be useless for that purpose
... I would be interested in feedback from others

Ted: while some individuals from OEM I have talked to defer on the political/legal topic, agreement on common data model and the other areas we are exploring in this task force are important for data marketplace regardless of pending legislation. we have had some data people from OEM involved attend on occasion, need to engage them more
... prods Benjamin

Benjamin: I am not so much focused on data marketplace, ontologies and providing solutions
... digital clone or shadow is a name for what we are doing, representation for a vehicle
... focused on modeling a vehicle

Ted: some parting words, TPAC is an opportunity for us to draw from the Auto WG who we would like more active and the broader W3C community. Please keep that in mind and see who you can draw in on automotive ontologies, policy languages, consent capture, permissions etc while you are there. Look at the unconference session ideas for any that might align with our interests, propose your own. relay in email any you feel of interest as I might know individuals who are organizing those

[adjourned]

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: 2018/10/04 16:27:24 $