Project Meeting 2025.04.08

Agenda

• Telecommuting design options

Technical

• Comparisons of SANDAG production and SANDAG example model
  • There are minor differences between
  • There is no significant differences that suggest taking telecommuting in a different direction
• Telecommuting design options
  • Presentation can be found here: https://drive.google.com/drive/folders/14v7yFqGrczmrFR7ltrsVWgRgEQiBUnCh
  • Current ActivitySim models
    • Telecommute frequency model, which is then used as covariate in downstream models
  • Work completed to date
    • Empirical analysis of survey data
    • Software enhancements: added Telecommute status model
  • Design considerations
    • Represent "hybrid" work patterns
    • Represent multiple work activities and locations
    • Describe in-home work activity schedules
    • Model sequencing
  • Options for representing "hybrid" work patterns
    • Expand telecommute status model
      • Replace telecommute status with Telecommute Arrangement
      • Alternatives: In-home work activities only; in-home + out-of-home activities; Out-of-home activities onl
    • Add "also out-of-home work activity" model
    • Expand coordinated daily activity pattern model
      • Add w/ in-home vs w/in-home patterns to existing DAP options
    • Use the existing mandatory tour frequency model - considered not an option based on previous discussions
    • Hybrid
      • Pro: Allow telecommuters to engage in an out-of-home work tour
        • According to PSRC, make up significant percentage of telecommuters
      • Con: Would need to select some arbitrary observed duration for an in-home work activity to "count" for this discrete choice; Need to re-estimate mandatory tour scheduling for "hybrid" workers
  • Represent multiple work activities and locations
    • Expand telecommute status model
      • Pros: Generalization of, and therefore superior to, "hybrid" suggestion; Explicitly represents multiple work activities and locations
      • Cons: Need to set a threshold for duration of "in-home" work activity; No data on more than one in-home activity; Re-estimate mandatory tour scheduling for those with >= 1 in-home work activity
  • Scheduling At-home Work Activities Options
    • Add in-home work activity duration model
    • Add in-home work activity scheduling (i.e., start time and duration) model
    • Add joint strategic work day arrangement model
      • Don't have data to pursue this option
    • Use the existing mandatory tour scheduling model
    • Pros: Software to explicitly represent at-home work activities; Sets the stage for more consistent and realistic itineraries
    • Cons: No data on in-home start and end times; Need to decide which to schedule first, in-home or out-of-home; Need to re-estimate mandatory tour scheduling
  • Recommended Solution: Multiple Work + Scheduling
    • Telecommute arrangement model: Addresses hybrid and multiple work activity
    • In-home Work Activity Scheduling
      • Software gives users the following options: 1. Duration Only. No blocking of time windows for other activities; 2. Start time and duration. Block resulting window from other activities
    • Duration outcome can inform downstream models
    • CDAP options: whether Telecommute arrangement model happens before, after, or in-conjunction with CDAP model
      • Recommend: CDAP then Telecommute Arrangement
  • Next Steps
    • April 17 - Submit draft recommendations memo
    • April 24 - Follow-up discussion on design, leading to a decision
  • Discussion
    • New Gallop Pole: Remote capable jobs and employees; Would be represented in the telecommute frequency model and the work from home model
      • SANDAG uses 2-digit NAICS code as variables in the model
    • None of the CDAP options are that ideal
• Thursday's meeting is cancelled