Skip to content

Project Meeting 2023.03.07

Michelle Bina edited this page Mar 8, 2023 · 5 revisions

Agenda

Key Decisions / Action Items

  • WSP to contact a limited number of interviewees (4-6) and report back to the group in 2 weeks
  • RSG/WSP to present BayDAG changes on Thursday
  • BayDAG changes to be merged into ActivitySim with no need for additional code review, prior to Jeff's work on ORCA

Meeting Notes

Follow-up on Roadmap

  • Interviews likely to be iterative, so focusing on the initial outreach now
  • Interviewees (initial outreach), propose the following priorities:
    • Consortium members and AMPO (Caitlin)
      • Vendors
      • Ben Stabler is a unique case
      • Also propose to include Bentley, Caliper and Inro
      • Benefits to talking to vendors include:
        • Understanding their long-term plans for integration and/or competition (ask for a summary of what they are doing now (in terms of activity-based modeling)
      • Questions that we would have for the consortium would be different from vendors, for example:
        • Don’t ask about ActivitySim management
        • Do ask things about data exchange/schema, etc.
      • Bench consultant developers
      • Jeff, Joel, David, etc.
      • Providing input in a separate reach out
      • Others – Elizabeth Sall, modelers in Victoria AUS, Vladimir (MAG)
    • Other comments
      • Don’t worry about academics
      • Next priority would be agencies that we don’t currently engage with (Marty Milkovits would be a good person to interview)
  • Next Steps
    • Revise contact list for initial interviews (reduce number of vendors and add a couple of non-consortium agencies)
    • Identify some open-ended and some specific prompt questions
    • First round of limited interviews
      • Purpose of limited initial interview is to prioritize topics of interest/content – somewhere between completely open-ended but have some specific/targeted topics to address for each interviewee but not a script
      • Start with just 4-6, focusing on consortium members and one non-member, to test this process.
      • Report back to the consortium in two weeks to discuss what’s working and what’s not, what are some things learned, then go back to the rest of the initial interviews.
      • The process could change/evolve after these first few interviews.
    • Second round of outreach could be a survey and then can be more open-ended in soliciting feedback and can get to a bigger, more specific questions/subjects that are specific to them
  • What does roadmap mean to us? Is it focusing on software, or what is critical/features prioritizes? Or is it the consortium vehicle? What is the overall objective?
    • Answer: All of the above. Speaks to philosophy of product management
      • What does the customer want? – speaks to the features
      • What can the engineers do? – speaks to the software
      • What can the consortium fund? – speaks to the consortium vehicle and future

BayDAG

  • SANDAG and MTC models supported by RSG and WSP have added new functionality into various components
  • Accumulated changes in the SANDAG fork and is ready to be merged
  • Key things
    1. Important for consortium to understand all the changes and accept them (most have already been presented, for example: joint-tour model extension)
    2. David and Sijia have already independently reviewed code changes and have tested the code. This suggests the review process should be minimal. It was proposed to not do another round of reviews, incorporating the ORCA changes into that pull request/review.
  • If no one is asking for another review of the BayDAG changes, we can accept merge now and then Jeff can do the ORCA changes afterward. This is the most efficient path forward (in terms of schedule and resources).
  • Agreed to not do a code review of the changes and just to merge.
  • David and Sijia to give tour of the changes on Thursday.
Clone this wiki locally