Next steps in describing possible effects of automated driving on traffic flow

Although at times traffic flow can be complicated to understand and predict, our current understanding is nevertheless pretty good and allows predictions of current and future situations to be made. However, this is about to change with the emergence of (partially) automated vehicles1on roads. This emergence will change the way people drive (speed and acceleration choices, following behaviour, lane change behaviour) and will change the dynamics present in traffic flow and with that the ability of traffic to move effectively. The first generation of partially automated vehicles is already on the road in the form of Adaptive Cruise Control (ACC) and Lane Keeping Assistance (LKA), as the so-called SAE level 1 vehicle automation (SAE 2014). Practical and technical tests are already widespread on higher levels of automation, with their mass-produced introduction expected in advance of 2020 (Merat et al. 2014). According to Shladover (2016) fully automated (SAE level 5) vehicles will not be here until 2075 at the earliest. At this point, no statement has been made on the point if these changes are going to be positive or negative. The reason for this is that there are too many uncertainties in relation to the effects that automated vehicles will have on traffic. Many enthusiasts have made incredible claims that automated vehicles will solve many of our traffic problems and can have an overwhelming positive effect on mobility, traffic flow and safety. However, these claims are rarely substantiated and are often based on questionable assumptions regarding the future. Furthermore, even if these claims are correct, they are applicable to a situation that lies many decades in the future, i.e. 100% penetration rate, and do not consider the route leading towards this future (the so called transition period). The main consensus within the traffic and mobility community is that traffic flow will initially worsen with the introduction of automated vehicles (Le Vine, Zolfaghari, and Polak 2015, SBD and HERE, 2016). This has a lot to do with conservative settings of automated vehicles, unfamiliarity with automated vehicles from both drivers and other road users, initial teething problems with the technology and a limited ability and penetration for cooperation between vehicles (Shladover, Su, and Lu 2012). Furthermore, there are predictions that traffic volume may eventually increase due to a steep uptake in automated vehicles (Wadud, MacKenzie, and Leiby 2016). However, again there is much uncertainty in relation to these statements, even if they are well substantiated Next steps in describing possible effects of automated driving on traffic flow. Available from: https://www.researchgate.net/publication/312377387_Next_steps_in_describing_possible_effects_of_automated_driving_on_traffic_flow?channel=doi&linkId=587c8ba108ae9a860fed444b&showFulltext=true [accessed Jul 18 2018]