In 2017, the German ethics commission for automated and connected driving released 20 ethical guidelines for autonomous vehicles. It is now up to the research and industrial sectors to enhance the development of autonomous vehicles based on such guidelines. In the current state of the art, we find studies on how ethical theories can be integrated. To the best of the authors’ knowledge, no framework for motion planning has yet been published which allows for the true implementation of any practical ethical policies. This paper makes four contributions: Firstly, we briefly present the state of the art based on recent works concerning unavoidable accidents of autonomous vehicles (AVs) and identify further need for research. While most of the research focuses on decision strategies in moral dilemmas or crash optimization, we aim to develop an ethical trajectory planning for all situations on public roads. Secondly, we discuss several ethical theories and argue for the adoption of the theory “ethics of risk.” Thirdly, we propose a new framework for trajectory planning, with uncertainties and an assessment of risks. In this framework, we transform ethical specifications into mathematical equations and thus create the basis for the programming of an ethical trajectory. We present a risk cost function for trajectory planning that considers minimization of the overall risk, priority for the worst-off and equal treatment of people. Finally, we build a connection between the widely discussed trolley problem and our proposed framework.

Autonomous vehicles are being developed to make road traffic safer in the future. The time when autonomous vehicles are actually safe enough to be used in real traffic is a current subject of discussion between industry, science, and society. In our work, we propose a new approach to the risk assessment of autonomous vehicles based on risk-benefit analysis, as it is already established in other areas, such as the registration of pharmaceuticals. In this context, we address the question of socially acceptable risk for mobility and investigate this concept as a decision-making criterion in trajectory planning. We make the first attempt to quantify an accepted risk by comparing autonomous vehicles with other types of mobility while taking into account the ethical and psychological effects important to the acceptance of autonomous vehicles. We show how an accepted risk contributes to the transparent decision-making of autonomous vehicles at the maneuver level. Finally, we present a method for considering accepted risk in trajectory planning. The evaluation of this algorithm in a simulation of 2,000 scenarios reveals that lower risk thresholds can actually reduce risks in trajectory planning. The code used in this research is publicly available as open-source software: https://github.com/TUMFTM/EthicalTrajectoryPlanning

Reliably predicting the motion of contestant vehicles surrounding an autonomous racecar is crucial for effective and performant ego-motion planning. Although highly expressive, deep neural networks are black-box models, making their usage challenging in this safety-critical applications of autonomous racing. On the other hand, physics-based models provide high safety guarantees for the predicted trajectory but lack accuracy. The method presented in this paper targets this trade-off. We introduce a method to predict the trajectories of opposing racecars with deep neural networks considering physical constraints to restrict the output and to improve its feasibility.We report the method’s performance against an LSTM-based encoder-decoder architecture on data acquired from multi-agent racing simulations. The proposed method outperforms the baseline model in prediction accuracy and robustness. Still, it fulfills quality guarantees of smoothness and consistency of the predicted trajectory and prevents out-of-track predictions. Thus, a robust real-world application of the model with high prediction accuracy is proven. The presented model was deployed on the racecar of the Technical University of Munich for the Indy Autonomous Challenge 2021. The code used in this research is available as open-source software at www.github.com/TUMFTM/MixNet.

Self-driving vehicles (SDVs) will need to make decisions that carry ethical dimensions and are of normative significance. For example, by choosing a specific trajectory, they determine how risks are distributed among traffic participants. Accordingly, policymakers, standardization organizations and scholars have conceptualized what (shall) constitute(s) ethical decision-making for SDVs. Eventually, these conceptualizations must be converted into specific system requirements to ensure proper technical implementation. Therefore, this article aims to translate critical requirements recently formulated in scholarly work, existing standards, regulatory drafts and guidelines into an explicit five-step ethical decision model for SDVs during hazardous situations. This model states a precise sequence of steps, indicates the guiding ethical principles that inform each step and points out a list of terms that demand further investigation and technical specification. By integrating ethical, legal and engineering considerations, we aim to contribute to the scholarly debate on computational ethics (particularly in autonomous driving) while offering practitioners in the automotive sector a decision-making process for SDVs that is technically viable, legally permissible, ethically grounded and adaptable to societal values. In the future, assessing the actual impact, effectiveness and admissibility of implementing the here sketched theories, terms and the overall decision process requires an empirical evaluation and testing of the overall decision-making model.