Network Change Guide Wishlist

[JoeJimFlood]

This discussion is for expressing desires for the Network Change scenario, including specifics on the scenario to test and what metrics to demonstrate the calculation of. It is proposed to implement a change to the transit network as that is a common contemporary travel model application. Suggestions from the April 7 consortium meeting included leveraging the transit scenario used in testing the initial EET implementation. Some proposed metrics to demonstrate the calculation of include:

• Tour mode share
• Travel times
• Jobs within a transit ride of a given time

Please submit any comments before the April 21 consortium meeting.

[sumitbindra]

Network Editing Scenario

Problem Statement

1. (TransLink preference) The addition of BRT service with 5-minute headways in the peak periods on a major arterial.

   • Which trip purposes show the strongest mode shift response?
   • What would be the time savings by trip mode and by purposes? How does reduction in travel time by mode impact mode shift for different trip purposes?
   • Are workers adjusting their work departure time based on availability of BRT services? (What if BRT services start at 4AM versus 7AM?)

2. If a new managed lane reduces peak travel time by 8 minutes for SOV users, does the model show induced demand in subsequent feedback iterations? How many iterations are needed before assignment and ActivitySim reach equilibrium?

Components Impacted (Edit based on modeling system for testing)

1. Mode Choice Response – Adding BRT or a managed lane should shift mode shares, but the question is whether the shift is reasonable and traceable. The guide should walk through how to read mode choice coefficients, how to calculate expected utility changes from the skim improvement, and how to sanity-check that the model's response is proportional to the level of service improvement.

2. Time of Day Choice – Network changes often have peak vs. off-peak implications that ripple into time-of-day choice. A managed lane that only operates during peak hours, for example, should be shifting some tours earlier or later. This is a subtle effect that senior modelers need to understand how to look for and explain to stakeholders.

3. Tour and Trip Assignment Feedback – Depending on whether the model runs with feedback iterations, a network improvement that reduces congestion can affect departure time and destination choice in subsequent iterations. The guide should address when feedback matters for a given scenario type and how many iterations are sufficient.

Outputs / Summary

1. Mode shift by trip purpose
2. Departure time impact
3. Congestion impact

[JoeJimFlood]

There's actually a BRT project in the SANDAG regional plan that I was thinking about suggesting using (it connects the suburb I lived in with a couple neighborhoods I probably would have gone to more had the line existed).

[xiex0055]

It is easy to set up the model to simulate the long-term travel effects of a permenant network change (e.g., the permenant closure of a river-crossing bridge). However, is there any guideline regarding how to set up the ActivitySim models to simulate the short-term effects of a temporary network change (e.g., the temporary closure of a river-crossing bridge for renovation)? Similar to the set-up in a trip-based model, we may want to freeze the long-term choices (e.g., workplace location, car ownership) and work related choices in such a scenario.

[JoeJimFlood]

@xiex0055 I could probably add a couple of paragraphs on this. I'd include instructions on saving the pipeline and setting the resume_after setting on the last model step to be frozen.

[amsamimi]

Hi @JoeJimFlood,
It could be valuable to test a Bike-and-Ride (BnR) scenario by adding BnR facilities at a subset of rail stations and introducing BnR as a mode in tour/trip mode choice. The main objective would be to evaluate how BnR affects transit catchment areas and station-level ridership, especially compared to walk and park-and-ride access.

[JoeJimFlood]

@amsamimi Thanks for the suggestion! It's really interesting and has me thinking about how it would be implemented. Unfortunately, I feel like that would be out of scope for the current task, though I think it would be a great application in the future (I suppose I could add something if there's budget left over).

[guyrousseau]

In the case of a implementing a change to a transit network, previous guidance from the Federal Transit Administration (FTA) for New Starts funding involved transit project sponsors using the SUMMIT model application to demonstrate that project benefits were being isolated to the actual project and that the variability associated with feedback loops and assignment weren’t being accumulated. This was done by using as many consistent inputs into the mode choice model as possible between the No Build and Build scenarios, and while the new guidance no longer requires the SUMMIT model, the lessons learned from its use still remain. In addition, whatever we propose should not be considered official guidance from FTA. It is the project sponsor’s responsibility to engage FTA in the methods that will be used for developing transit forecasts as part of a New Starts application.

[guyrousseau]

From a road network perspective, I would like to see an ActivitySim case study where we would implement a change to a road network. For instance, as an example, when introducing a new toll lane that is managed with pricing, the user will likely need to perform an iterative process in which ActivitySim is run to generate toll demands which are then run through a series of roadway assignments where the tolls are optimized. The new toll prices are then fed into another ActivitySim run to estimate new demands and the process continues until the demand and toll prices stabilize, since toll prices have an effect on toll eligible demand. Ideally, an ActivitySim toll optimization process would work towards an equilibrium between the generalized costs between the managed lanes and the general-purpose lanes. This optimization might be different from maximizing revenue. For managed lane corridor studies, it would be the user’s responsibility to determine the most appropriate toll rates for those studies. This process would be intended to assist modelers in finding a logical balance between travel time savings and tolls.