Trust, the firm belief in the reliability, truth, or ability of someone or something, underlies all social and economic relations and is central for the acceptance and adoption of autonomous vessels both by the maritime community and the general public. Trust requires explanations but is a much broader concept facilitating interaction among people and between people and technologies. Autonomous vessels are facilitated by artificial intelligence (AI), automating tasks previously performed by people, meaning that roles, responsibility, authority and decision making are delegated to data and algorithms. The need for trust, however, remains unchanged, but people become dependent as well as changed by these technologies. Simultaneously, multiple layers of interaction between people and technologies will likely continue to exist. People need valid explanations and causal reasoning for trusting critical, surprising or unexpected behaviour or decision making, also in the context of autonomous vessels, as incorrect behaviours or decisions can quickly translate into critical consequences. Such trust also depends on technical assurance processes where we emphasize that explanations can and need to play a role as valid evidence. We argue that the current methods for explaining AI are insufficient in providing trust in autonomous vessels and are too narrowly framed towards developers of AI. With the multiple points of interaction between people and autonomy, we argue that it is urgent to identify and mature explanation methods suitable for all types of interactions during development, assurance or operation. Explanations should be adapted to roles and responsibilities, and aspects such as context, cognitive skills, alertness, contextual knowledge, and time available to act by the user. We propose four types of explanations to suit the developer, assurance, end-user, and external explanation needs, which must be mapped out before the design such that trustworthy, interacting autonomous vessels can emerge.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.