Modelling the Impact of Automated Driving – Private AV Scenarios for Germany and the US

Modelling the Impact of Automated Driving – Private AV Scenarios for Germany and the US


Lars Kröger, German Aerospace Center Berlin, Tobias Kuhnimhof, German Aerospace Center Berlin, Stefan Trommer, German Aerospace Center Berlin


The paper presents projections for the impact of private autonomous vehicles on destination choice, mode choice and thus overall travel demand for Germany and the US by combining a vehicle technology diffusion model and a travel demand model.


Vehicle automation technology advances at rapid pace and the market entry of automated vehicles (AV) can be expected within the next years. Vehicle automation technology transitions gradually through different levels of automation (level 1 through level 4). However, substantial impact on travel choices only seem likely once drivers do not need to attend to the driving task anymore for most of a trip; i.e. drivers can take their β€œbrain off” and engage in other activities such as work or entertainment. This is likely to impact on travel choices such as destination and mode choice because drivers might be willing to spend more time in the car or because the car is more attractive relative to other modes. At the moment, the future outlook in terms of AV regulation does not include the prospect of AVs being allowed to move without a driver; i.e. there must be a driver on board able to take over the driving task. This prospect rules out autonomous shared vehicle systems (autonomous car sharing, autonomous ride sharing) to a large degree, making privately owned AVs a likely scenario.

The paper presents results from modelling travel behavior impacts of introducing AVs into the private car fleet. In order to model such a 2035 scenario, we combined a vehicle technology diffusion model and an aspatial travel demand model and applied this to Germany and the USA. Differentiating by passenger car segment, we introduce AVs among the newly registered vehicles from 2021 onward assuming an s-shaped market-take-up until 2035. By then, 50% of the new vehicles and 25% of the passenger car fleet are projected to be AVs. Again differentiating by segment and age, the AVs can be found among specific driver groups. In addition we assume that AVs are owned by mobility impaired travelers who did not have the option to drive previously. Subsequently, we use a travel demand model consisting of trip generation, distance choice and mode choice to forecast travel by different traveler groups and by car availability (no car, conventional car, AV). For modelling the impact of AVs compared to conventional cars, we reduced access/egress times due to quicker parking / valet parking and we reduced values of car travel time savings for travelers with AVs.

While the model results overall conform to expectation the impact of AVs on travel behavior are not large: There is a ~5% increase in VMT for both Germany and the USA, resulting from somewhat longer trips combined with slight modal shifts from other modes towards the car. These results have important implications: If the regulatory framework for AVs is such that a private AV scenario is the most likely development, AVs are not likely to revolutionize travel. AVs will change travel behavior – but their impact might be marginal compared to other external factors.


Association for European Transport