Generally, the input data for DynusT can be classified into four categories: Demand, Supply, Calibration, and Validation.
Demand - the demand category data is primarily the time-vary origin-destination (O-D) matrices, that is generally put into an hourly interval (can be shortened to intervals in minutes). The trip roster data are typically for either scenario comparisons or for activity-based model (ABM) integration. In the case of scenario comparison, one may want to use the trip roster produced in the baseline case and load that to all compared scenarios. This would ensure that the demand remain identical for all compared scenarios. This technique is needed to capture the performance difference due to scenarios while fixing the demand. For ABM integration, the trip roster is the direct output of ABM. Loading these trips directly to DynusT without converting back to OD matrices is one step toward ensuring the consistency between ABM and DTA.
Supply - the supply side data usually includes all the network (node, link, zone) attributes, as well as all network controls such as signals and U-turn flags.
Calibration - link counts and speed profiles are needed for the Stage 1 and Stage 2 calibration.
Validation - the data used for model validation usually include screen line or turning movement counts, TMC travel time/speeds that can be provided by 3rd party vendors like INRIX or HERE or specific probe vehicles commissioned to collect GPS data for the project.