Telecommute

Work-in-progress

Telecommute Model Plan

In response to Phase 6a Task 3 Telecommute Model in Cooperation with SEMCOG

Background from Wu

• On SANDAG’s ABM wiki site, please find SANDAG ABM2+ Enhancements to support 2021 RTP (2020). The telework methodology section starts on page 10.
• Additionally, we have done some interesting telework testing by varying telework rates, the report and results can be found here: SANDAG ABM2+ Sensitivity Testing Report (2020). The telework section starts at page 74.
• You are welcome to check out other parts of our wiki site. The sensitivity testing reports might be useful to some of you.

Proposed Design

Joel's proposed design presentation for SEMCOG based on the previous SANDAG work. In summary:

• Informed by SANDAG and MAG models
• Add a person worker telecommute frequency model
• Add this telecommute freq variable to cdap, indiv non-man tour freq, non-mand tour stop freq
• Add work from home extension to work location choice model
• Only apply telecommute freq model to workers with usual workplace outside home

Key outstanding issues

• Should we include person occupation and work industry since these are key determinants of telecommuting but not in every synthetic population?
  • Probably since the code should work in either case - with or without these additional person table columns --> yes
• Should we revise the usual work location model to include a work from home, work away from home extension?
  • Probably since we want to differentiate usually working from home from telecommuting as a replacement for travel --> yes
  • However, we plan to add a person work_from_home submodel after work location choice that makes use of the already output work_location_choice_mode_choice_logsum for accessibility and overrides the work_zone field
  • This is cleaner since it separately the work from home model from the work location model, which has benefits for shadow pricing, estimation, model design flexibility, and software implementation flexibility
• Can we make it any easier for the user to do what-if analysis? Maybe some kind of smart constant calculator for target telecommute shares or something like that? --> we'll add this to the issues for a potential future improvement