Models of Human Decision-Making as Tools for Estimating and Optimizing Impacts of Vehicle Automation

Abstract

With the development of increasingly automated vehicles (AVs) comes the increasingly difficult 2 challenge of comprehensively validating these for acceptable, and ideally beneficial, impacts on the transport system. There is a growing consensus that virtual testing, where simulated AVs are 4 deployed in simulated traffic, will be key for cost-effective testing and optimisation. The least 5 mature model components in such simulations are those generating the behaviour of human agents 6 in or around the AVs. In this paper, human models and virtual testing applications are presented for 7 two example scenarios: (i) a human pedestrian deciding whether to cross a street in front of an approaching automated vehicle, with or without external human-machine interface elements, and (ii) an AV handing over control to a human driver in a critical rear-end situation. These scenarios have received much recent research attention, yet simulation-ready human behaviour models are lacking. They are discussed here in the context of existing models of perceptual decision-making, situational awareness, and traffic interactions. It is argued that the human behaviour in question might be usefully conceptualised as a number of interrelated decision processes, not all of which are necessarily directly associated with externally observable behaviour. The results show that models based on this type of framework can reproduce qualitative patterns of behaviour reported in the literature for the two addressed scenarios, and it is demonstrated how computer simulations based on the models, once these have been properly validated, could allow prediction and optimisation of AV impacts on traffic flow and traffic safety.

Multi-agent traffic simulations to estimate the impact of automated technologies on safety

“…For example, we modeled driver reaction in its interaction with a pedestrian by incorporating reaction times randomly extracted from an empirical distribution. However, recent studies on a controlled single car-to-pedestrian interaction suggest that driver reaction depends on the relative kinematics between the agents involved and hence more advanced models that capture this interaction may be needed (Markkula et al 2018). A combination of well-controlled and validated single-event approaches, with applications to the multi-agent approach, may lead to an improvement of both the understanding of the underlying interaction mechanisms between agents and the relevance of these mechanisms in terms of overall impact on safety in wider and more complex environments.…”

Why did the human cross the road?

“…Rasouli et al [2017] introduced datasets of interactions between pedestrians and human-driven vehicles. Most studies agree that pedestrian's crossing decision depends mostly on vehicle dynamics which can be summarized using the time to collision (TTC) parameter [Markkula et al 2018]. [Schneemann and Gohl 2016] show that vehicle speed, is the most determinant factor of pedestrian's decision process.…”

Applicability of risky decision-making theory to understand drivers' behaviour during transitions of control in vehicle automation

“…It must be noted that the rate of evidence, or "drift", is accumulated differently for every person, which is also influenced by a number of factors. In the field of vehicle automation, Markkula et al (2018) have demonstrated how to apply decision-making models based on evidenceaccumulation to explain, for example, what information drivers use to decide how to resume control from vehicle automation to avoid an incoming forward collision. Bounded rationality models, first defined by Simon (1972), which holds that humans can make decisions based on the information available to them.…”

“…Currently, there are only a few studies that highlight the possibility of such a link (c.f. Markkula et al, 2018). In this work, we consider how theoretical models for risky decisionmaking can be used to study drivers' transition of control in automation by observing their visual sampling behaviour during different stages of the take over process.…”