Arman: Arman Aygen, digital
transformation and cybersecurity in mobility and been involved
with US DOT, SAE, ISO etc
... interested in improving collaboration and security
Adnan: Adnan Beknan, BMW Munich R&D with focus on data architecture
Clemens: Clemens Portele from
Interactive Instruments in Germany, involved in Spatial Data
management and very active in OGC
... route API and data model work at OGC will be my main
focus
Hongki: Hongki Cha from ETRI,
first time joining a W3C Auto call
... interest in developing with the draft spec under
discussion
Benkamin: Benjamin Klotz, been working on formal modeling/ontology of vehicle data
Gunnar: Gunnar Anderson, technical lead from GENIVI
Kryztof: Kryztof J@@ from UC Santa Barbara and involved in various ontologies including trajectories and traffic prediction
Linda: Linda van den Brink, I work for Geonovum in the Netherlands and heavily involved in Spatial Data standards, author of SDW Best Practices, chair of SDWIG and OGC
MagnusF: Magnus Feuer, chief
artchitect for telemetics JLR
... also looking at next gen entertainment systems, one of
starting participants in VSS at GENIVI and looking to expand
signals and interaction including RPC
Mark: Mark Fox, University of Toronto director of SmartCities there, working on city ontologies
Marty: W3C Business Development
Megan: Megan Katsumi, University of Toronto PostDoc Fellow working with Mark on City data
Tom: come from ITS 204 and
IOC/IEC SmartCities
... most interested in services than data
Ken: Ken Vaughn, chair of WG1
articture for ISO Intelligent Transportation Systems ITS
TC204
... working on ontologies and looking forward to coordinate
with SmartCities, W3C etc
... want to coordinate data as much as possible
Ulf: Ulf Bjorkengren senior connectivity
strategist at Geotab
... interested in API design and all around connectivity in
automotive field
Wonsuk: Wonsuk Lee ETRI,
government sponsored research institute of South Korea and been
part of this group since 2015 and editor of VISS, had worked on
prototype
... interested in API and framework to support applications and
services outside the vehicle
... also of implementation of this group's specs as an open
source project
... last year's project focuses were elsewhere
Sanjeev: Sanjeev BA with Samsung,
based in Korea and had previously collaborated with GENIVI, W3C
and OCF
... currently involved in Gen2 client development
ThomasS: Bosch, VISS Vehicle application plugin platform for 3rd parties as has been presented at GENIVI in the past
Ted: Ted Guild, I lead the W3C Automotive activity
Raphael: EURECOM in Sophia Antipolis and been involved in W3C many years, chaired several groups and worked with Benjamin, Daniel and Adnan on VSSo
Philippe: Philippe Robin GENIVI lead and working with Gunnar on CCS
<raphael> I was co-chair of the Media Fragments URI working group but that was 10 years ago :-)
Karen: Karen Meyers, support outreach for W3C
MagnusG: Magnus Gunnarson
Mitsubishi Electric, we're Tier 1 suppliers to OEM and was at
the first meeting in Rome a long time ago
... interested in connectivity and data model
Peter: Peter Winzell software engineer at Volvo cars and WG chair, interest in vehicle signals
Ted reviews agenda
Ted: we want to reach out and
include different industries and focus areas
... telematics, regulators, insurance cos, intelligent
transportation systems, smart cities
... have a much broader picture for data sets
... On screen I'm showing the charter and areas of work
... we will continue to coordinate with GENIVI on VSS
... exposing vehicle applications
... There is an ontology that Raphael and Benjamin were discussing
with BMW
... BG will also work outside charter...take ontology and take
real data
... have a graph server running and some canned questions and
queries
... that answer some real world canned questions with real
data
Ted: to show the power of this
outside of the cloud
... important to show the power of this and for adoption
... on policy level, there will be some challenges in the
current client
... there is Gen2 server applications; potential work in the
WG
... makes sense with the graph project to have that in the
business group
... we want to work on application best practices
... applies to in-vehicle and cloud based applications
... make use of data; regarding permissions and access; what to
do with data
... various things as far as polling frequency
... not undermine control units
... not impact someone's anti-lock breaks
... cybersecurity best practices
... things we have done with DoT and others
... coordinate other group development surrounding data
usage
... some other work at W3c
... needs to be developed more clearly
... some are auto-specific, others more broad in how to use
data
... capture consent, things of that nature
... has been some work on that
... but needs to be split out on its own
... there is a mini-apps group at W3C
... a privacy group at W3C that we should work with moving
forward
... besides Gen2 server to promote this data model
... and service APIs
... more broadly
... make sure people are aware and not have duplication;
mapping
... try to get this out there
... make sure this work succeeds; enlist participants in doing
so; solicit where, how to do this
... we very much welcome ideas
... for how to go about that
... we will be discussing a nice topic
... major component of this business group is coordinating
around ontologies in the transportation space
... as far as cities, etc. and coordinate with researchers and
so forth
... another goal
... is to act as an incubator for ideas and tangents that come
off of the Auto Working Group's activities
... the high level bits
... especially the ontology coordination; will work with the
Spatial Data on the Web IG
... Linda is on this call as chair
... also Clemens working on this a
... I took over as staff contact with SPD WG; there is
coordination with OGC
... Initial chairs of the Auto and Transportation business
group
... will be Adnan Beknan, BMW and Glenn
Atkinson, Geotab
... People interested can do break-out
... use BG repository
... we use GitHub extensively for coordination
... If there is desire to have calls to further things along,
we can help to organize them
... We have mailing lists to be used for minutes and
announcements
... the group has decided to start using Slack
... which is a convenient way to stay connected; it's a real
time communication
... phone app
... hope people will use it going forward
... I've been talking quite a bit
... High level review of charter
... charters for business groups can evolve dynamically based
on interests
... if there are areas/interests that are missing, please speak
up
... if they are not there at the outset, we can evolve
this
... unlike the working groups that need AC and W3C review and
approval
... Do seek your feedback, if there are questions about the
scope, please speak up now
MagnusF: There is cross-over
<ted> MagnusF: there is cross over between VSS and Gen2
MagnusF: how do you see the BG and WG tracks interact with one another?
Ted: there will be
coordination
... make sure minutes are funneled through BG and Auto WG
mailing lists
... so people can follow
... don't expect everyone to be involved in all these areas,
but can follow at higher levels
... join calls, comment on issues in GitHub
... some conversations are better as GitHub issues; others
should be added
... if an open source implementation, feature, something wrong
in spec
... suggestions to bring up issues in GitHub
... we can also introduce by email
... if unsure, just pick one and get it out there
Magnus: Ok
Ted: Make sure there is funneling and cross-over
[Ted mentions coordinations]
Ted: things will be somewhat
scattered; a bit part is to pull some of these things
together
... high level questions, things that should be on the
additional scope, things you think we are missing?
... anything along those lines, please unmute
Peter: there are a lot of
different topics on there
... I think there needs to be some prioritization
... My concern there is too many topics
... any comment on that?
Ted: I agree
... we have a whole new cast of people coming on from the data
side
... some people have already focused in an area, like graph
data with Sangi; Megan and Mark on the ontology side
... I agree it's ambitious
... agree there is a lot there
... as far as the ontology side, there will be prioritization;
that is part of the agenda
... top priority is likely going to be routing ontology
... and some things may, if we don't get enough...good
cross-section on this call
... if there are people who should be part of the
conversations, please make introductions to me [and Marty and
Karen] about how to get involved
... if we cannot get them going, we will remove them from the
scope
... Is Glenn here?
... Adnan, would you like to say something?
Adnan: I agree we need to figure
out the main focus
... will align with Glenn
Ted: Peter is the chair of the
working group
... not too many changes over there; VSO spec work
... when further along it will go to WG, but will start off in
the BG
... scope for WG has not dramatically changed
... other than people starting to do demos
... for now
... it is a lot
Magnus: One question
... VSSO
... will main work take part in W3C BG or at GENIVI?
Ted: VSS is at GENIVI
... VSSO Eurocom is Raphael and Benjamin will put into the
repository
... hope to find out more from them this week
... it's using quite a few of the ontologies created by the
Spatial Data group
... ontologies belong more at W3C
... using those building blocks
... that's why it will be over here
... of course we have a great liaison and collaboration with
GENIVI
... a project from Gunner and Philippe overlaps with this
scope
... probably makes sense to do a brief intro now?
Gunner: Maybe better supported
with a few slides
... and provide context for those who are not familiar
... since we have people here who are not from core auto
industry
Ted: Do you want to share your screen?
Gunner: no, I don't have that right now
Ted: range of scope and different
aspects
... you will here from Gunner later this week about a project
in common
Gunner: it has been running for
over a year
... to correct that point
... we have started this with intention of aligning with
VSS
... about stated architectures
... we have done that work putting placeholders in place to
include W3C technologies and protocols
... expect W3C to complete those building blocks
... and build a bitter picture
... if needed for other non-W3C tech and protocols that have
been around a long time
... you are right this bigger scope for BG charter
... it does seem to overlap some of the thinking we have had
for quite a while
... in terms of taking charge of this work
... I don't see a problem that we won't be aligned
... it should be complementary, hopefully
... that is my quick intro
... work together to reduce fragmentation and to work together
to build these architectures
Ted: as a teaser
... I am broadcasting a slide
... we will hear more about that
... Bosch has a piece of this, too
... We should get you and Tom synched up
Gunner: we have had these discussions with Bosch about vehicle abstraction
TS: Sebastian has shared this image; is that what you are referring o?
Gunner: Christian and his
colleague
... we are working on
... trying to clarify for industry
... the big picture and what needs to be done
... we work with autostore and other orgs to put together these
complete systems both inside and outside the car
... want more time to distill that
... we have a bigger understanding
... in large part this is a renaming of what GENIVI has always
done
... to agree on standard interfaces with the industry
... the connected side is requiring standards there
... because of the infrastructure you are involved in
... there need to be common protocols and standards across the
vehicles
... put this into a combined view fairly soon
TS: From Bosch's perspective,
this is a high priority
... we want to standardize; I think everyone does
Gunner: that is what we have been working on for the last ten years
Ted: I'm thinking...
... under VSS update; regarding our splitting up tomorrow
... Gunner and Thomas, wondering the best time for you both to
give your respective presentations
... That would be 9pm for you
TS: yes, I'm good with that
Gunner: yes, 9pm timing works
Ted: I think that's where that fits logically
Gunner: we see this as a big
picture
... and every problem needs to land back on VSS in some
sense
... and the kind of standardization that is needed; there is a
relationship for sure
Ted: Trying to accomplish
... before we take a break
... Megan is going to do an ontology 101 for us
... Useful to describe
... helps to look at our in-vehicle ambitions
... how we want to get things out there
... but make sure people understand once we have data in the
cloud, how do you structure it, represent it
... high level scope for the task force, track 2
... the transportation coordination committee
... what is it about
... Megan, can you share your screen now?
... I have added your slides to the agenda page
... People can open the PowerPoint and follow locally
Megan: this is meant to be a high
level introduction about what we mean about ontologies
... we're from University of Toronto Enterprise Integration
Lab
... more recently focusing on the area of SmartCities, we're
here from the School of Cities which Mark is Director
... transdisciplinary involving engagement and
partnerships
... we have hosted some webinars and will be more in the
future, we have a satellite in India as well
... brief introduction on topic of ontologies, provide a
foundation
Slides (Powerpoint) Slides (PDF)
Megan: it is not my intent to
start a debate on definitions but provide our usage
... and a spectrum
... when we talk about ontologies we are looking at core
concept and properties and provide enough structure for
semantic integration, automated deduction
... this is definition of a city resident and how an ontology
would be useful representation of the individual
... definition of a resident varies by city
... we need a mechanism to distinguish between these
meanings
... this shows how to formalize these nuanced semantics and
include shared properties across them
... it is done in a logical language, we are focused on OWL and
RDF(S)
... our main criteria are formalisms and contain axioms for
metrics of the terms
... just because it is in an OWL file doesn't make it an
ontology, there should be formal semantics too
... this diagram illustrates broader range of things considered
ontologies and highlight toward the right what we
consider
... the others are useful but want to be clear about the types
of references we are talking about
... finally to give a concrete idea for ontology for people who
haven't seen them before, what to recognize in a formal
ontology
... enable microtheories
[diagram doesn't cooperate in Libre office for MagnusF who is presenting slides nor Ted scribe]
Megan goes over diagram about properties of a resident
Megan: here we go into more
detailed axioms, example being owner of a business in Toronto.
this shows a foramization of that statement in OWL
... enables automated reasoning and checking
... micro-theory: applications where you build more complex
rules even in a different language for more sophisticated
reasoning, this one about probability of different
activities
MagnusF: from your example if certain criteria isn't met for a given definition is it still used
Megan: that is one of the uses
Ted: Thank you very much
... That is helpful
... goes into more depth
... than what some people worry about on a day-to-day
basis
... people can add to it
... there are ways to flow from VSS to VSSO
... ways to use ontologies
... people from vehicle side
... there will be different external data sets to bring into
the vehicle from the Spatial Data folks
... will be useful for people to start thinking about
that
... limited edge computing
... challenge to bring that volumne of data and the computing
power to consume
... there are some ways
... that semantic rich data gets 'dumbed down' into lesser
formats
... if ther is advice from those who are more versed in
semantic web technologies
... some high-level advice to give regarding those needs
... ways for these two groups to coordinating moving
forward
... how to figure out some specifics
... maybe not have to do it now; but keep it in mind
... for what we expose for data out there and what we may want
to bring back into the car as well
... just get people starting to think about it
... and realize there are people who can help us
Gunner: there is logic, reasoning and deductive reasoning; not sure how much is going to be run in the vehicle
Ted: a car may not make a routing
decision and calculate a myriad of data
... but may be given a representation of a route
... when we get to vehicle infrastructure information
... data from vehicle to cloud
Gunner: it is a distribution
mechanism
... normalization
... we probably want to run a Harvard model
... may have to download enough data
... to look at decision making
... this is a huge new field; take my comments with a grain of
salt
Ulf: I have not thought about it
before
... where computational would be
... it could be a fact
... of where you have most of the data needed for this
... is most of data is in the cloud
... better to move that data from the vehicle up to the cloud
and do the computation
... or possibly the other way
... to reduce the connectivity load
: Do we have a concrete use case of vehicle-cloud integration?
Peter: Looking at supply chain,
real timeish
... building bottom up
... with vehicles
... and what they have and what they are and using customer
description
MagnusF: that is a very important use case
ThomasS: Bosch is looking at such use cases with software as service
ThomasS: there is a huge push for definite use cases like this, businesses as a service and something we're working on
Adnan: we are currently aligning toward VSS but metadata in the cloud, where it will be in a half an hour's time
MagnusF: offboarding data one thing, bringing it in and processing in-vehicle such a use case
Ken: we have a bunch of use cases and architectures between objects with US DOT and want to develop ontology with uses in mind
Ted: I believe it was
Ken who just spoke
...Pull in the two
views
...you can tell us
what your visioning should be on the edge
...in the vehicle or
at the traffic light
...this is a great
topic
...a bunch of TBD
placeholders
...for different
sessions
...I want to have
architecture and vision discussions
...Bosch is saying
yeah, we want to do this
...but may not be
clear where it belongs
...hearing from
Magnus and myself, there are limits on the car
architectures
...I will put in a
topic for later in the week
...encourage
...I almost want to
upend this; it's also aligned with what Thomas and Gunner plan
to present
...maybe put this
topic into the agenda earlier in the week
...if any of the
people I just mentioned, or others feel strongly about the
architecture topic
...and if you have
limited availability this week, please let me know
...as I shift a
couple things
...Planned to do
intro of the Transportation Ontology Community
...will provide link
for second track
...go from there a
bit
...As far as
additions to the agenda
...different
architectures, who's doing what/where, I think it should be
tomorrow
...Is anyone not
available tomorrow?
@@: I am available all
day tomorrow
Ted: we have time to
carve out for this important conversation later
...The transportation
ontology coordination community
...keep it brief
...we know that we
want to contribute VSSO from the automotive side
...some things
Benjamin mentioned before that need to be there
...We want to look at
SSN extensions and see how that applies to us
...VSSO will be
contributing as far as vehicle data out to the cloud
...we held a workshop
in Sept 2019
...several people on
this call were there
...Mark and Megan,
Ken; OGC representatives
...Clemens will
discuss routing API and ontology
...which is of high
interest to all concerned; both in-vehicle and in-cloud
...important to avoid
stepping on each others' toes and avoid duplication of
efforts
...look at design
principals
...look at spatial
data best practices
...VSSO does a
modular approach
...a CG does this in
schema.org
...uses SSN and
Sosa
...others can
represent it far better than I
...we'll discuss this
later
...We want to get
input from spatial data folks
...real goal is to
catalogue what is going on; who has what
...take design
criteria mind
...encourage changes
to it
...there are key core
ontologies that can be built upon
...route,
observations
...are examples
...several of these
efforts
...and see what is a
good core ontology
...and build upon it
for specific needs and use cases
...traffic adjusted;
observations made during a route
...original goal to
kick this off
...during these
meetings
...he's not on
today's call
...RJ who was at the
Sept. workshop from
Volpe DoT
...we are trying to
get people on the same page
...see from vehicle
side what we are producing
...and separate
topic
...what information
we want to bring into car v decision made in cloud
...then make
instructions
...back down as a
result of those actions
...I'll stop
there
...Also see if others
like Ken, Megan or Mark
...anyone who has
been part of these conversations, would like to jump in
...and give their
ideas
...this is an initial
idea; they can be refined
...have more detail
when we get into the breakout
...take a break after
this
..and then split into
two different tracks
Ken: I am happy to present on the architecture
Ken: happy to present on
architecture
...not sure who was
at the workshop in Sept.
@@@: It would be useful for me
Ted: go for it [Ken]
Ken presents screen, architecture for US DOT ARC-IT.org
Ken: this is a complete
architecture, particular 42010, we have four different views
and working with those ISO committees on types of views
... who owns, operates, maintains equipment, functional view on
real time operations
... a field of all use of information technology for
transportation, more focused on public sector
transportation
... traffic signal control
... functions to be performed, processing in the field
... communications examines interactions between objects
... [use cases]
[diagram with different objects - pedestrians, vehicles, lights...]
Ken: we started defining what we
feel should be computed where as was being discussed
earlier
... focus has been on public sector side but could be applied
to private as well
... this shows the information flows between objects, you can
get details by clicking on them
... including involved standards
... or if there is a gap
... nice thing about this architecture is it provides 140
different use cases
... some of this of interest to public safety, port
services...
... how connected vehicle security is maintained
... this is a reference architecture designed for local
regional implementations with their needs in mind
... this is a US DOT effort but will need to vary by
state
... individual project level architecture can be abstracted out
of this
... the tool, Vizio, allows you to modify and custom flows as
needed
... would be easy for people on this call and provide their
architectural views, different than we may
ThomasS: are you familiar with C@@ (OEM and SmartCity data) project?
Ken: we currently have these four
views, planning a fifth which would be information flow
specific to see how ontology relates to component
... while this originated in US, it has international content
already
... we've been working with JP AU and EU, bring in their data.
it will take time to fully flush out but working on making it
more universal
[break]
Peter: Ulf will give
overview of gen2 roadmap
...Vehicle Signal
Specification the data model being used . Maintained by
Genivi
... VSS not complete yet,
but choosen as recommended data model
... VIWI was another
model considered
... support for full
subscribe.
Peter: http methods and addressing the VSS tree is
Peter: Jason Web token used in access restriction
Peter: VSS data model changes, separation of node type and data type
Peter: type nodes in the same branches as other nodes
Peter: Instantiation support for VSS added
Peter: Pending access restriction not complete
Peter: Dynamic registration...data tree not static thru its lifetime
Peter: In its early stages
Peter: Versioning of VSS also needed for VISS
Peter: Gunnar (Genivi) has made a proposal for the above
Peter: Queries, Ulf to present new functionality tomorrow
Peter: 2 different documents: CORE and TRANSPORT
Peter: Transport: https://www.w3.org/auto/wg/wiki/Gen2
https://raw.githack.com/w3c/automotive/gh-pages/spec/Gen2_Transport.html
Core https://raw.githack.com/w3c/automotive/gh-pages/spec/Gen2_Core.html
Peter: Will probably not mandate data model, more recommend
Peter: reference to VSS (data model) needed
Peter: borrowed parts from VIWI
Peter: Needs an overview
Peter: Auth for the above two comments
Peter: Interface chapter needs to be rewritten since it is based on VIWI which will not be the recommended data model
Peter: Transport document
Peter: Some changes and tweaking of the document needed
Peter: Transport protocols (HTTPS, WEB SOCKETS) , PR now yet not merged
Peter: Formal chapters needs work
Peter: Out of scope, signal grouping, Distributed server, payload not JSON, other transports
Peter: Everything is still up for discussion
Peter: JSOM LD was up for discussion
Magnus Feuer: Where do we need help
Ulf: use functionality in GitHub projects
Ulf: Perhaps use Kanban
Magnus Feuer: Maybe next step
Magnus Feuer: We can use slack for coordination
Magnus Feuer: Guthub
Ulf: we need more editors
Magnus Feuer: A list of tasks would be useful
Ulf: Normative language needs to be added
Ulf: calls out for people to get an overview what's there in the specs
Magnus Feuer: Using issues and slack
Magnus Feuer: Asks not taks
Thomas Sperecky: Layered discussion
Thomas Sperecky: Is layers concept part of gen2
Ulf : yes
Gunnar : Layeres useful for a number of things. VSS already defined for this
Gunnar: except for access control
Gunnar: not concluded yet
Ulf: confirms
Gunnar: private branch could be considered layered
Gunnar: private branch already defined in VSS
Gunnar: part of VSS definition
Ulf: The result of the layers modifications transparent to gen2
Ulf: Not needed to say anything in the gen2 documents.
Ulf: leave it to the vSS spec
Ulf: More as reference in the spec docs
Gunnar: the protocol do need the metada
Gunnar: add metadata . Gen2 do not know this
Gunnar: access control overlap
Gunnar: Naming , Gen2 not a good name
Ulf: gen2 just working name
Ulf: very opiniated
Ulf: no consensus
Gunnar: Add to to do list
Ulf: conclude GitHub, slack and a list of asks
Ulf: please join and help
Magnus Feuer: RPC discussion
Ted: Should it go in to gen2
Ted: The needs for RPC
Great
MagnusF: viss are way to publish
signals data to one or more subscribers whereas rpc is about
sending a reply, functional call
... we may want to standardize specific calls
[slides RPC scope and requirements]
Magnus: what I want is agreement
on the problem statement, what we are trying to solve and its
requirements?
... what is it used for BYOD, call types
... robustness expectation, arguments, service discovery - at
build or run time
... this is a vehicle centric view but that needs to
changfe
... RPC vs pubsub, RPC for initiating change in system as a
basic strategy
... security - how to authenticate/authorize and how does this
align with VSS security
... signals and RPC will be used in tight conjunction from
cloud
... we can draw lessons from our RVI (Remote Vehicle Interface)
work at GENIVI
... what tooling is available, their strengths - FrancaIDL,
YAML...
... who needs this and what are their needs?
... that is the outline
Gunnar: Franca and YAML have different levels of abstraction
MagnusF: question is what is the appropriate level of abstraction needed
ThomasS: we have spent a fair amount of internal discussions at Bosch on modeling this, or model of models
Gunnar: seems like we missed a big opportunity for collaboration with Bosch a year ago
MagnusF: we have people working on this as well
MagnusF iterates some various use cases, diagnostics, vehicle interactions
MagnusF: also for external
services and smartcity integration
... another would be applications deployed within vehicle in a
carefully controlled sandbox environment, may want to increase
their access over time
... how to control the interactions with outside world
... I see signals and RPC intertwined
Ulf: this seems to be a big jump in scope to Gen2?
MagnusF: could be part of larger roadmap and later phase
ThomasS: this sounds good here
MagnusF: this is part of similar
architectures being pushed by Tier 1s
... do we want to start to standardize some of these
further
Ulf: instead of being built into Gen2, it could be above it and influence underlying capabilities
MagnusF: I am open to that, whatever works best
Ulf: my initial reaction is on top
Peter: do you mean a separate specification?
Ulf: yes
... this would utilize Gen2
Gunnar: how could that work?
Ulf: clearly provides view into signals, at the end it boils down to them
MagnusF: if you look at SOME
IP
... we are working in Gen2 on REST replies and can be handled
with it
Gunnar: yes, request/reply can provide RPC
MagnusF: it is extending concept
Gunnar: I agree
... clarifying - actuate and change signals. it would be rather
odd to have a signal that writes 20% of window open compared to
a move window function
... another plug for WAMP
MagnusG: wouldn't integrating RPC
into the spec mean any OEM could define any functions they want
internally
... one of the benefits of the W3C spec is it commonizes data
access and do not want to loose that
MagnusF: depends on how wide a
scope we want to take on, could start with navigation with
destination, routing engine interaction
... or look at standardizing on/off board but not sure which
level to approach it from