Skip to content

Project Meeting 2024.10.10

rcopperman edited this page Oct 22, 2024 · 1 revision

Agenda

  • Admin updates
  • Telecommuting Explicit Representation

General Admin

  • Admin updates
    • Not much to report on new Admin home; Mark has sent Joe an update on his findings regarding AMPO.
    • No update on payments and MOAs
    • Cancel meeting on 10/15

Technical

  • Telecommuting Explicit Representation
    • WSP outlined a list of 5 Tasks being undertaken by WSP for SANDAG.
      • Survey Analysis
      • Initial Sensitivity Testing
      • Software Changes
      • SANDAG Model Configuration
      • Final Sensitivity Testing
    • Today, WSP is reporting on what will be done for the first two tasks.
    • Asserting that Telecommuting is cosmetically unattractive as implemented in ActivitySim
      • Persons simulated on sim day do not engage in a work activity and don’t show up as such in the simulation. Just given a home activity pattern; confounds telecommuters with those who don’t go to work for other reasons.
    • Is it behaviorally unattractive in that it has an undesirable impact?
      • Yes. Currently treated in ActivitySim as a daily phenomenon, but this ignores partial day telecommuting. There is no in-home activity because it does not create a tour.
    • Behaviors related to telecommuting:
      • Work tour activity scheduling
        • Want to simulate work activities that occur at home – in survey data?
        • Do workers have the same schedule as those going to work in survey data?
      • Want to determine through the survey analysis if this explicit modeling of telecommuting leads to different work start/end times, durations. Want to know whether there is no behavioral impact or if the survey is not up to the task.
      • Want to know if telecommuting frequency or status has an effect on:
        • Availability for school escorting
        • High propensity for non-work tours and trips.
    • Guy asked about the scope reference to "commute impedance", can this be further defined?
      • Joe: "commute impedance" is related to difficulty getting to a workplace, which might make someone more likely to telecommute.
    • Guy also asked about use of industry codes. Joe said there is a coarse representation by four groups.
      • Joel clarified: The initial model used 2016 SANDAG survey data, which did not capture occupation. RSG tested some industry specific effects. Survey was only about not going into work, not partial work at home. He cautioned that any survey data summarized to compare telecommuters to non-telecommuters needs to be aware of day of week differences—lumpy in the expanded data which is not expanded by DOW. Too many Tue/Thu will miss people who work remotely one day a week (usually Wed, Fri).
    • Plan is to test with the SANDAG model and compare with the survey results
      • Survey questions related to telecommuting frequency are not great. Do half days count as one day of frequency?
      • Need to consider Telecommute status on the survey day – number of hours
      • Need work start times and durations. Surveys do not include this for at-home work activities.
      • If we have telecommute frequency and status, then we can tell if status is important above and beyond frequency.
        • Impact on nonwork activities – shop, rec
        • Impact on escorting children
      • Telecommute status – need to establish some rule, maybe a duration threshold
        • Joel: where you draw the line is important
      • For chauffer pattern, model already considers whether the person has a work tour when linking students with workers.
      • Guy: how do you account for presence or absence of children in the households?
        • Dave: we look for children up front
          • Considering Ohio and SANDAG surveys. Ohio are pre- and post-COVID era
          • Joe: we are more interested in post-COVID era surveys
          • Dave: work activity scheduling will not be very robust for work telecommuting. But should be able to get at other trip purpose scheduling.
    • Next Steps:
      • WSP is now evaluating data and will recommend which surveys to use and how
      • Sensitivity testing
    • Joel asked whether the scope for this work is available. Joe said he would share it.
    • Guy: Will this impact estimation data bundle work?
    • Dave: There is no explicit relationship assumed. They are trying to determine whether re-estimation is warranted.
    • Joe: Scope does not include estimation work, potential for future.
Clone this wiki locally