Web Conference 2025.08.12 Curb - openmobilityfoundation/curb-data-specification GitHub Wiki

CDS Working Group

Agenda

CDS 1.1 Draft Review - Milestone - Release Draft (leave comments) - all changes - browsable dev branch

  • Welcome (5 mins)
  • Announcements (5 mins)
  • Presentation: INRIX and SFMTA - Curb Objects (10 mins)
  • CDS 1.1 Draft Review (40 mins)
    • Vehicle properties and computer vision fields #196 - change
    • Add color options to Policy in the Curbs API #195 - change
    • Adding Enforcement Attributes to Demand Data #179 - change 1 2
    • Add Payment Channel and Method to Curb Events #158 - change

Organizers

  • Hosts: Michael Schnuerle, OMF
  • Note Taker: Michael Schnuerle, OMF
  • Facilitator: Michael Schnuerle, OMF
  • Outreach: Michael Schnuerle, OMF

Action Items and Decisions

  • Review and comment on all proposed changes by Aug 22.
    • Vehicle properties and computer vision fields #196 - change
    • Add color options to Policy in the Curbs API #195 - change
    • Adding Enforcement Attributes to Demand Data #179 - change 1 2
    • Add Payment Channel and Method to Curb Events #158 - change

Minutes

Notes

INRIX and SFTMA presentation

  • Showing details of CDS implementation
  • Introducing details of drafted Right of Way specification

CDS 1.1 Release Plan and dates

Vehicle properties and computer vision fields #196 - change

  • Vehicle types align to MDS
  • track lanes that events take place in
  • Do we need both detection and recognition confidence?
  • Vehicle fields will move to its own object in 2.0, but for now all inline
  • 1 to 100 on confidence level

Add color options to Policy in the Curbs API #195 - change

  • New policy color object in curbs API
  • policy_color had digital typo
  • Add dot_dash to primary border type
  • Color policy itself is optional

Adding Enforcement Attributes to Demand Data #179 - change 1 2

  • 4 new types of events - detection, violation, citation, end
  • Eg length of double parking violation
  • Metrics updates too
  • Enforcement data object to hold values
  • enforcement_id is used to tie together event types
  • status could be another field -
  • location and time comes through in normal Event data fields
  • Uniform renaming suggestions - need start and end events for consistency
  • Do all starts need an end, vice versa?
  • typically require is confusing, should it be required or not? Make it clear.
  • vehicle detected is not necessarily a violation, more for occupancy
  • rename vehicle_violation to violation_start
  • time violation was observed, vs when it occurred
  • Solving for both real time and non real time - can append historic data to events when known.
  • Append old violation data into Events to compare horizontally in one place
  • Allows for any kind of event entry type/source
  • what other fields are needed in Enforcement object?
  • Consensus around requiring a violation end? End should require a start...
  • Might not need an end, because it could be a moment in time, and time is not relevant.
  • Danko from Passport to draft update.
  • Grace period - included in local violation rules, and no violation unless it's beyond grace period. Do you need to record the grace period at all pre-violation? Likely not.

Add Payment Channel and Method to Curb Events #158 - change

  • Is it right to have event and then payment method and channel?
  • What else is needed, or consolidated?
  • Add channel: telephone - automated phone tree backend
  • multiple payment vendors... which is captured?
  • data source operator - is the payment source, like an app or meter company
  • data source device - is the hardware source, like a meter
  • other issue opened about vendors in Metrics, issue 190 for 2.0

September 9 we will review draft of CDS 1.1

Chat

  • 00:06:03 Michael Schnuerle (OMF): Agenda: https://github.com/openmobilityfoundation/curb-data-specification/wiki/Web-Conference-2025.08.12-Curb

  • 00:09:46 Brian Hamlin - Seattle DOT: Love when the three legged stool diagram comes out

  • 00:10:15 Sam Brenner: Reacted to "Love when the three ..." with ‼️

  • 00:10:43 Michael Schnuerle (OMF): Reacted to "Love when the three ..." with 3️⃣

  • 00:11:17 Christopher Shelley | City of Philadelphia: What is this "maintain" word?

  • 00:13:36 Michael Schnuerle (OMF): Reacted to "What is this "mainta..." with 🫣

  • 00:16:01 Kenya Wheeler (SFMTA): Reacted to "Love when the three ..." with 3️⃣

  • 00:16:06 Brian Hamlin - Seattle DOT: Reacted to "What is this "mainta..." with 🫣

  • 00:16:24 Michael Schnuerle (OMF): Two-three minutes left, thanks ⏰

  • 00:17:14 Graham Rossmore (LADOT Parking): Kenya, can you share these slides with me afterwards? Would love to take a deeper look

  • 00:18:36 Kenya Wheeler (SFMTA): Hi Graham, happy to share these slides with you. They should also be in the meeting notes on GitHub as well @Michael Schnuerle (OMF)

  • 00:18:45 Michael Schnuerle (OMF): Reacted to "Hi Graham, happy to ..." with πŸ‘

  • 00:19:33 Brian Hamlin - Seattle DOT: Reacted to "Hi Graham, happy to ..." with πŸ‘

  • 00:21:01 Brian Hamlin - Seattle DOT: Thanks!

  • 00:22:30 Michael Schwartz, INRIX: Would love to dive deeper with folks on this topic. We will be at ITS and presenting with OMF on this and the broader Road Rules implementation at our booth. Please be in touch if you will be there and would like to discuss more

  • 00:22:32 Michael Schnuerle (OMF): Release Plan https://github.com/openmobilityfoundation/curb-data-specification/wiki/Release-1.1.0

  • 00:24:35 Jacob Malleau (CurbIQ): Would be great to hear what additional attributes you feel Curb Objects needs to help translate assets <-> curb zones/policies. Maybe a discussion for a future call.

  • There were some attributes we left out when making a baseline for Curb Objects, but definitely some additional info could make that change management work easier (on CurbIQ's end, we have flagged items like arrow direction, connecting signs, etc.)

  • 00:25:23 Graham Rossmore (LADOT Parking): Reacted to "Hi Graham, happy to ..." with πŸ‘

  • 00:27:59 Kenya Wheeler (SFMTA): Under vehicle properties, what is the standard that we are using to measuring confidence levels for vehicle type or paint? Is there an industry standard or is this based on "professional judgement"?

  • {I can leave this on GitHub as well]

  • 00:30:14 Brian Hamlin - Seattle DOT: Reacted to "Would be great to he..." with πŸ‘πŸ»

  • 00:30:18 Kenya Wheeler (SFMTA): Reacted to "Would be great to he..." with πŸ‘πŸΎ

  • 00:30:19 Brian Hamlin - Seattle DOT: Reacted to "Under vehicle proper..." with πŸ‘πŸ»

  • 00:30:33 Jerad Weiner - San Francisco Public Works: Reacted to "Under vehicle proper..." with πŸ‘πŸ»

  • 00:43:59 Jerad Weiner - San Francisco Public Works: There are other non-vehicle violations that might exist in the curb space. Temporary construction staging (permitted or unpermitted) in the ROW for example

  • 00:44:38 Brian Hamlin - Seattle DOT: Will this format allow for non real-time enforcement event tracking? Meaning, will a city be able to translate existing citation data (separate table provided by police department) into enforcement events even if they weren't done via ALPR?

  • 00:45:24 Michael Danko (Passport): Replying to "Will this format all..."

  • You could derive citation issuance events from issued citations.

  • 00:46:15 Brian Hamlin - Seattle DOT: Reacted to "You could derive cit..." with πŸ‘πŸ»

  • 00:46:17 Kenya Wheeler (SFMTA): Reacted to "Will this format all..." with πŸ‘πŸΎ

  • 00:46:18 Jacob Malleau (CurbIQ): Replying to "Will this format all..."

  • Agreed. You likely wouldn't use the violation_Start and violation_end in this case, but rather just citation_given

  • 00:46:24 Kenya Wheeler (SFMTA): Reacted to "There are other non-..." with πŸ‘πŸΎ

  • 00:46:27 Brian Hamlin - Seattle DOT: Reacted to "Agreed. You likely w..." with πŸ‘πŸ»

  • 00:48:59 Jerad Weiner - San Francisco Public Works: Replying to "There are other non-..."

  • There might be another space for this topic or maybe out of scope but didn't want get overly focused on vehicles as being the only users of curb space.

  • 00:49:10 Kenya Wheeler (SFMTA): Replying to "There are other non-..."

  • There are other non-vehicle violations that might exist in the curb space. Temporary construction staging (permitted or unpermitted) in the ROW for example

  • Those are important to capture. A permitted use of the ROW could be considered a non-violation, while an unpermitted use would be eligible for a violation (at least as I understand this question).

  • 00:49:21 Michael Schnuerle (OMF): Payment proposal https://github.com/populus-ai/curb-data-specification/tree/event-payment-type/events#payment-channel

  • 00:51:37 Jerad Weiner - San Francisco Public Works: Replying to "There are other non-..."

  • Correct, also thinking about semi-permanent curb space occupiers like parklets or other types of curb space usage beyond vehicles that gets enforced by departments other than the transit agency or DOT.

  • 00:52:07 Jerad Weiner - San Francisco Public Works: Reacted to "There are other non-..." with πŸ‘

  • 00:53:03 Michael Schwartz, INRIX: Replying to "Would be great to he..."

  • Would be great to hear what additional attributes you feel Curb Objects needs to help translate assets <-> curb zones/policies. Maybe a discussion for a future call.

  • There were some attributes we left out when making a baseline for Curb Objects, but definitely some additional info could make that change management work easier (on CurbIQ's end, we have flagged items like arrow direction, connecting signs, etc.)

  • @Jacob Malleau (CurbIQ) great questions. Since we got information from SFMTA assets in a different format it is hard to say what else would be needed if we had gotten them from Objects data alone. I will bring this up with our technical team and see what additional object attributes we left in to make this linking possible. Definitely a topic for a future discussion

  • 00:53:42 Jerad Weiner - San Francisco Public Works: Reacted to "Would be great to he..." with πŸ‘

  • 00:55:34 Jacob Malleau (CurbIQ): Reacted to "Would be great to he..." with πŸ‘

  • 00:58:12 Kenya Wheeler (SFMTA): Reacted to "Would be great to he..." with πŸ‘πŸΎ

  • 01:00:07 John Lundstrom: We define grace periods of 5 minutes to allow systems to synchronize.

  • 01:00:22 Michael Danko (Passport): I think the proposed change supports that. Detecting a vehicle is before/during the grace period and detecting a violation is post-grace period

  • 01:00:23 John Lundstrom: But that is purely for citation issuance.

  • 01:00:33 Wei Cheng: Reacted to "Agreed. You likely w…" with πŸ‘πŸ»

  • 01:00:34 John Lundstrom: Reacted to I think the proposed... with "πŸ‘"

  • 01:00:34 Kenya Wheeler (SFMTA): Reacted to "I think the proposed..." with πŸ‘πŸΎ

  • 01:00:43 Wei Cheng: Reacted to "Will this format all…" with πŸ‘πŸΎ

  • 01:01:00 Brian Hamlin - Seattle DOT: Need to jump. Thanks everyone for discussion today.

  • 01:01:49 Stefanny Perez (CurbIQ): Hits from LPR can be the flag/ alert and the other data can come from the violations data when the citation is issued and confirmed by an officer (after the grace period)

  • 01:02:07 Jerad Weiner - San Francisco Public Works: Thank you!

  • 01:02:10 John Lundstrom: Thank you!

  • 01:02:15 Kenya Wheeler (SFMTA): Thanks all!