Project Meeting 2021.05.13

Project admin

RSG working to wrap up phase 6a and team looking forward to working with the expanded project team
Welcome MWCOG and Ohio DOT!
- I'll add Mark and Rebekah to the standing meeting invites, which are optional
- There's also a partners only meeting once a month
- Plan to discuss website/marketing at next partners only call
- Let's add the MWCOG and Ohio DOT logos to the ActivitySim website
MetCouncil starting their ActivitySim development project
- In phase 1, will setup an existing example with MetCouncil skims, syn pop, and land use
- Phase 2 will be a year later and will do local calibration/estimation/improvements
- Will share progress with team and solicit feedback for ideas
It is exciting to see development being guided more and more by partner efforts
Let's see how we can structure phase 7 to best align with this idea
GitHub and Travis accounts transferred to AMPO
MTC migrating to EMME and ActivitySim
- Phase 1 is the EMME migration and Phase 2 is ActivitySim
- Also currently looking at visualization tools
- And planning a 2 zone versus 3 zone system comparison task which several folks are interested in
- Already had a CityPhi demo too
ARC interested in issues #414 and #409

Further discuss comments on task 3, telecommute model
- Discussed where to put the work from home model and how it relates to what accessibilities are available to use
- SANDAG model is inside usual work location
- SANDAG model doesn't use the usual work location destination choice mode choice logsum, but a very similar measure computed ahead of time by SANDAG's accessibility calculator
- Asim idea was to put after work location in order to use the destination choice mode choice logsum, but this is funny sequencing
- The Asim aggregate accessibilities calculator from TM1 doesn't have the same accessibility measure from SANDAG so we probably need to use something similar
- Let's review some of the implementations in practice - TM1, SANDAG, DaySim, and others
Joel working on task 4, transit pass ownership model use case examples
I'm now testing the chunking improvements for task 7, usability enhancements
- Improvements won't necessary make the model run faster, but it should help with figuring out the optimal chunksize, which can help make more efficient use of available resources, which can help free up more resources, which can then be utilized to help reduce runtimes
- I'll test a 128gb and 256gb example for example to help us understand the new feature
Will do a release when these features are ready, this will be version 1.0