Airport Model Improvements: Results Analysis Framework

[DavidOry]

For the previous central mobility hub planning work, we created a mini data pipeline to extract the results from the airport model and combine them with the outcomes from other scenarios, and then rendered key results using Tableau dashboards.

How do we want to extract and show results with ABM3? Are existing procedures in place for summarizing airport model results across scenarios? Do we have all the requirements from the planning team re: what is needed? Do we want to revive the Tableau summaries or move to Power BI or SimWrapper or something else?

cc: @rcu, @AshishKuls

[rickcurry]

@bhargavasana

[AshishKuls]

There is also a data pipeline and Tableau dashboard for comparing survey and model outputs (ABM2+). We would want to do some similar analysis to evaluate the current ASIM implementation of airport model.

[bhargavasana]

Currently, we have a power BI dashboard to compare ABM3 model run scenarios. It automatically pulls data from outputs written to Azure datalake (https://sandag.github.io/ABM/design/report/report.html). It shows basic mode choice for airport model and can be extended to show more summaries. @JYSD1 has been working on a new calibration viz to compare model and survey numbers (https://github.com/SANDAG/CalibViz). I am hoping to develop this also as part of this project.
cc: @cundo92 @kelvinnguyenn

[cundo92]

I agree with @bhargavasana. Ideally no new platform / software would be introduced and all summaries would be handled by our existing reporting procedures that work off of uploaded scenario data onto azure. From there, summaries / reporting would be produced in databricks and/or Power BI.