Sustainability of Autonomous Vehicles: An Agent-based Simulation of the Private Passenger Sector

Initiatives such as the European Green Deal establish mandatory objectives for climate neutrality by 2050. To realize this, predominant sectors like the transportation industry necessitate substantial improvements in emission efficiency, as transportation accounts for 20% of global CO2 emissions. Remarkably, 41 % of that are attributed to passenger transport alone. Amidst technological progression, shared autonomous vehicles (SAV) are projected to be a vital instrument in reducing greenhouse gas emissions in the private passenger sector. Hence, this study investigates the potential sustainability advantages of SAV introduction to the private passenger car sector.

We draw upon an agent-based simulation model to avoid building upon theoretic populations and generic simulation approaches, not appropriately allowing to derive realistic SAV simulations. Agent-based simulations account for individual agent optimisation and are posed to be especially applicable to model traffic. More precisely, we build upon a calibrated model which incorporates real world commuter and travel statistics of Berlin. As current research frequently uses outdated travel data, this study generates a projection of three levels of travel demand of the wider area of Berlin in 2050. ... mehrAdditionally, we identify the need for more research on sustainability effects considering multiple levels of potential SAV introduction, as AV and SAV adoption is associated with high uncertainty. Regulatory interventions posed to be a solution to steer SAV adoption effectively. Therefore, we introduce three levels of SAV-exclusive car-based traffic zones in our simulation scenarios. The three level of zones range from the inner city of Berlin to the entire simulation, including Berlin and Brandenburg. Lastly, we identify the need for more comprehensive sustainability analyses, focusing on more than single parameters such as tailpipe emissions. Consequently, we compare driving-related emissions and energy consumption as well as the total expected life-cycle greenhouse gas impact of all SAV introduction and demand forecasting scenarios.

Our findings reveal that SAV introduction increases the total passenger travel duration up to 62.1% and the passenger travel distance of up to 15.2% due to added wait time and detours. This effect is particularly noticeable in large SAV-exclusive zones. However, occupancy rates increase simultaneously, causing total vehicles kilometers to stay consistent. We observe an initial rise of 0.8% to 2.8% in vehicle kilometers considering an unchanged population, while smaller SAV-exclusive zones see the highest increase. In turn, when including our travel demand forecast which accounts for increased population size and travel density, SAV introduction reduces total vehicle kilometers by 0.5% to 3.6%, related to higher SAV occupancy. Lastly, accounting for SAV-related efficiency increases, we conclude savings in total life-cycle CO2 emissions ranging from 0.4% to 9.6% and energy consumption ranging from 1.5% to 12.2% across all scenarios. When combined with a fully electric SAV fleet, the potential for emission reduction increases to 59.0%, and for energy consumption reduction to 74.7%.