Abstract

Robotic and Artificial Intelligence (AI) systems within industrial settings are growing and changing rapidly. The trends of increasing intelligent automation and connectivity lead to transformations and challenges, especially in the realms of ethics and responsibility. Currently, robot and AI ethics lack mechanisms of reinforcement. Ethical guidelines and recommendations often are ineffective, and the question remains, of how ethics can be integrated successfully and easily applicable. This paper is based on the results of an empirical study on the perception, distribution, and governance of responsibility and ethics within robotic engineering in industrial robotics. Within a research design of Grounded Theory and Actor-Network Theory this paper roots in the findings of a set of 29 qualitative items; seven expert interviews, four expert talks, and 18 narrative interviews. The three main results are the following: Firstly, robot and AI systems contain a hybridity of actors, forming a multi-level robot actor network, of actants on the individual, corporate, collective, and im-material levels. Secondly, this complexity leads to ethical challenges within the robot actor-network. Thirdly, the data-based solutions of concrete actions for ethical engineering and engineering ethics inform a multi-level-framework for integrating ethics in robotics. The proposed Ethics in Robotics and AI, EROS framework bridges abstract normative codes and guidelines with concrete actions and visions from the domain of robotic and AI engineering itself. Finally, it serves as a starting point for providing a future kit of tools, and mechanisms for all levels within the robot actor-network, to positively integrate ethics in an application-oriented manner.