Agent-Based Model for Southern California Estimated a 10 Percent Increase in Vehicle Miles Traveled and Emissions with the Adoption of Connected and Automated Vehicles.

Several Travel Demand Management Strategies Were Tested for Southern California Using Activity-Based Demand Model to Evaluate Their Impacts on Reducing the VMT and Emission Growth Disbenefits Caused by Connected and Automated Vehicles.

Date Posted

Connected automated vehicle impacts in Southern California part-II: VMT, emissions, and equity

Summary Information

Connected and Automated Vehicle (CAV) technologies have the potential to significantly impact people’s travel behaviors and the performance of transportation systems. This study investigated the impacts of CAVs from various aspects, including Vehicle Miles Traveled (VMT), emissions, and transportation equity in Southern California. A comprehensive model was developed by incorporating the supply-side improvement of CAVs, a modified activity-based demand model (ABM) supported by survey data, and a multi-class highway assignment model. Discovering some disbenefits associated with the CAVs based on the model results in this study, the researchers in this study designed a set of demand management strategies and policies based on the Southern California Association of Governments’ (SCAG) travel demand management (TDM) strategy toolbox, aiming at resolving the issues caused by CAVs, and then assessed the effectiveness of reducing excessive VMT and emissions and resolving equity concerns.


The SCAG ABM simulates the travel behaviors of individuals in Southern California by considering seven layers of complexity, which generates a synthetic population for the SCAG region, predicts individuals' medium-term and long-term choices, simulates interactions among household members, determines trip departure time and mode choices, and assigns activity-based travel demand to the physical network to estimate vehicle miles traveled and other performance metrics, to support transportation planning decision-making. 

In this study, a stated-preference survey was designed and distributed in six counties in Southern California to gather people’s willingness to use CAV and the travel behavior changes associated with the CAV deployment. Based on the survey data, multiple models were recalibrated to incorporate the CAV-associated choice behavior changes of CAV users in the SCAG ABM framework. Finally, the following TDM scenarios were tested based on the strategies highlighted in the TDM toolbox summarized in SCAG TDM strategic plan report: telework (Scenario 1, S1), transit fare subsidy or free transit (S2), parking pricing (S3), and auto trade-in (S4). Multiple performance indicators, including VMT, vehicle hours traveled (VHT), traffic speed, mode share, trip length, and equity were used. 


  • The simulation results showed that VMT and emissions would increase both by 10 percent under the CAV Base scenario, and that CAVs could worsen travel equity across income groups. 
  • After deploying TDM strategies, although the VMT and emissions were still increasing in S1 to S4 compared to the SCAG model, all scenarios indicated reduced VMT, and emissions compared with the CAV Base model.
    • S1 resulted in a 7.9 percent increase in VMT and eight percent increase in CO2 emissions. This means S1 resulted in less disbenefits (about 2 percent points less than the CAV-only case without the telework demand management strategy). 
    • S2 resulted in a nine percent increase in VMT and 9.2 percent increase in CO2 emissions. This means S2 resulted in less disbenefits (about 1 percent point less than the CAV-only case without the transit fare subsidy demand management strategy).
    • S3 resulted in a 8.6 percent increase in VMT and 8.8 percent increase in CO2 emissions. This means S3 resulted in less disbenefits (about 1.4 percent points less than the CAV-only case without the parking pricing demand management strategy).
    • S4 resulted in a 3.0 percent increase in VMT and 2.7 percent increase in CO2 emissions. This means S4 resulted in less disbenefits (about 7.0 percent points less than the CAV-only case without the auto trade-in demand management strategy).
  • Equity evaluation results showed that the telework policy (S1) offered greater equity improvement in terms of household trip numbers and travel distance than the other scenarios, possibly due to the more significant reduction of commute trips in middle and upper-income groups.
Results Type
Deployment Locations