Prior to the peer review, SFCTA staff identified a list of issues and topics of interest to guide the discussion during the peer review. These issues were presented to the panel members on the day of the review. The specific questions and panel's responses pertaining to the topics of interest are presented in Section 6.
In the shorter term, SFCTA is interested in assessing current state of calibration and the strategies that may be feasible to implement within the scope of the current project. The Authority would also like to consider the following approaches in the long-term if they offer promise for DTA and model improvement:
There are bus-only lanes in the CBD which can be used by other vehicles if they are making a right turn. SFCTA is interested in appropriately modeling this phenomenon. Restricting the bus-only lanes to only buses reduces the capacity of links and results in gridlocks throughout the CBD. To eliminate gridlocks, SFCTA has currently removed restrictions on bus-only lanes and allows their use by all traffic. An idea to split links containing such lanes into a bus-only part and a part with right-turn bay is being explored as an improved method for modeling this. Other complications include violation of these bus-only lanes by other vehicles and the fact that taxis could use these lanes. SFCTA had issues with underestimation of traffic flow on major arterials and overestimation on local and collector roads. To overcome this, the free-flow time on local and collectors roads has been doubled. SFCTA is interested in other approaches to deal with issue of matching observed traffic flow patterns.
SFCTA would like to identify best generalized cost formulations and optimal DTA run settings for the model.
Having collected its own data for estimating traffic flow parameters for the DTA model, SFCTA has a strong interest in finding out other data collection methods that might help improve the accuracy of the parameters. SFCTA used data from three sources - PeMS, SFMTA speed surveys, and SFCTA traffic flow surveys to help answer the following questions:
SFCTA also noted the following challenges it faced while conducting the traffic flow surveys:
The Authority is highly interested in standard validation practices used for DTA models.
Integration strategies for demand (SF-CHAMP) and supply (DTA microsimulation) models are a primary focus issue for SFCTA. The goal is to achieve better consistency between supply & demand models to avoid unreasonable gridlock. As mentioned earlier, demand matrices from SF-CHAMP are directly loaded into the DTA model without any "adjustments" or "alterations". SFCTA would like to improve the level of consistency by creating better LOS information for demand models and creating better demand information for DTA models. The following are some specific topics of interest in this regard.
Other practical considerations include the stability of results across scenarios resulting from the disaggregation options specified above and convergence strategies. The level of disaggregation could also be dependent on the policies being considered.
In addition to all travel demand information, DTA could also use user heterogeneity that is available from the demand model. In SF-CHAMP, each traveler has their own value of time which could be incorporated into DTA through additional user classes. This would require additional time and restructuring DTA software.
One of the goals of the DTA project was to provide a good starting point to other MPOs and agencies which may be embarking on a similar endeavor. SFCTA is very keen on knowing how the knowledge gained through this project may be effectively disseminated through the modeling community. The Authority is interested identifying various components of this project that may be more useful such as data processes, code bases, user guides and other documents.
SFCTA is also interested in finding out if there are any research questions that could be answered during the course of this project.