Automated cars meet human drivers: responsible human-robot coordination and the ethics of mixed traffic

Abstract

In this paper, we discuss the ethics of automated driving. More specifically, we discuss responsible human-robot coordination within mixed traffic: i.e. traffic involving both automated cars and conventional human-driven cars. We do three main things. First, we explain key differences in robotic and human agency and expectation-forming mechanisms that are likely to give rise to compatibility-problems in mixed traffic, which may lead to crashes and accidents. Second, we identify three possible solution-strategies for achieving better human-robot coordination within mixed traffic. Third, we identify important ethical challenges raised by each of these three possible strategies for achieving optimized human-robot cordination in this domain. Among other things, we argue that we should not just explore ways of making robotic driving more like human driving. Rather, we ought also to take seriously potential ways (e.g. technological means) of making human driving more like robotic driving. Nor should we assume that complete automation is always the ideal to aim for; in some traffic-situations, the best results may be achieved through human-robot collaboration. Ultimately, our main aim in this paper is to argue that the new field of the ethics of automated driving needs take seriously the ethics of mixed traffic and responsible human-robot coordination.

On the Impact of the Rules on Autonomous Drive Learning

“…This way the solution of a moral dilemma should be calculated based on a given set of rules or other mechanisms-although the exact practical details and, most importantly, their corresponding implications, are unclear. The problem of autonomous vehicles regulation is particularly relevant in mixed-traffic scenarios, as stated by Nyholm and Smids [25] and Kirkpatrick [26], as human drivers may behave in unpredictable ways to the machines. This problem could be mitigated by providing human drivers with more technological devices to help them drive more similar to robotic drivers, but mixed traffic ethics certainly introduce much deeper and more difficult problems [25].…”

The ethics of crashes with self‐driving cars: A roadmap, II

“…This type of “mixed traffic” is perhaps the most realistic situation to anticipate for the coming years ahead. On the supposition that self‐driving cars would actually be safer than conventional cars, our suggestion was as follows: People have a duty to either switch over to the safer alternative, namely, autonomous cars, or to use or accept added safety precautions when using the less safe alternative, namely, conventional cars (Nyholm & Smids, forthcoming).…”

“…The latter – namely, seeking means for making people drive more like robots – might seem to constrain people's freedom and their personal autonomy on the road. There is a need for more ethical debate about how to handle these sorts of problems with human–robot coordination in this domain, as well as in other domains into which robots are increasingly being introduced (Nyholm & Smids, forthcoming).…”

A Conceptual Control System Description of Cooperative and Automated Driving in Mixed Urban Traffic With Meaningful Human Control for Design and Evaluation

“…traffic lights). There are also questions about gaps in control of CAVs [2,3] and moral choices by CAVs [4][5][6] that are present for highway traffic, and these are all more potent and amplified in the urban traffic environment [7]. With certain aspects of CAV control being questionable from a purely operational and human acceptance perspective [8,9], the concept of Meaningful Human Control (MHC) has been adopted for Cooperative and Automated Driving (CAD) to explicitly include human moral reasoning and ethical acceptability.…”