Designing for trustworthiness - training for trust - An overview of trust issues in human autonomy collaboration in maritime context

Aug 26, 2026

Trust is a central challenge for autonomous maritime operations, especially as intelligent technologies create more interactive, context-dependent human–technology communication. Research links human autonomy teaming to psychological mechanisms similar to human–human teaming. The study identifies trust issues in autonomous maritime use and connects them with trust and trustworthiness concepts through qualitative group interviews and a data-driven literature review. HUMANE workshops derive four meta-categories and outline design and training improvements, including calibration and explainable AI.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages9

Reading Time23 min

AccessFree to download

Abstract

Background and research interest

Method

Results and implications

Trust issue meta-categories

Design and training improvement areas

Introduction

Autonomy degrees and the maritime role of humans

RIA technology and sociotechnical complexity

Conceptual frameworks and trust in multi-level systems

Designing for trustworthiness, training for trust. An overview of trust issues in human autonomy collaboration in maritime context A. Hynnekleiv1*, M. Lützhöft1 1Department of Maritime Studies, Western Norway University of Applied Sciences *corresponding author Abstract - Background: Trust is recognised as a highly relevant issue for autonomous maritime operations. Increased agency of intelligent technologies enables interactions with humans to be bilateral, contextdependent and in result, more social. Research suggests that human autonomy teaming is a subject to similar psychological mechanisms as teaming between human actors. Research interest: Our study aims to identify trust issues linked to autonomous maritime operations and connect them with the existing concepts of trust and trustworthiness. Method: Qualitative analysis of group interviews and datadriven literature review. Results: The Human Autonomy Enable (HUMANE) workshops identified four meta-categories of trust issues, including operational trust, trustworthiness, organisational trust and social acceptance. Each category of trust issues is linked to existing concepts. The results point towards areas for improvement in maritime design and training: measuring perceived autonomy, application of explainable artificial intelligence, exploring linguistic and cultural aspects of communication, facilitating calibration of trust, supporting cooperation in instant teams, and responsible promoting of social acceptance. Keywords Ergoship 2021, maritime Human Factors, autonomous maritime operations, MASS, trust, trustworthiness, human autonomy teaming. Introduction The perspective of introducing autonomous maritime operations brings up a range of challenging questions regarding the human role in the sociotechnical system. In the regulatory scoping exercise, the International Maritime Organisation differentiated between four degrees of autonomy, including automated processes and decision support for seafarers (degree one), remote control of a ship with seafarers onboard (degree two), remote control of an unmanned ship (degree three) and a fully autonomous ship (degree four) (IMO, 2021) . At least in the first three degrees of autonomy suggest that humas maintain their key role in decision making. The fourth degree of autonomy is defined by the ability of an operating system to make decisions by itself, without human intervention. However, the degree of autonomy of a ship could change depending on conditions. The same vessel could operate as fully autonomous, but under certain conditions be controlled by an operator on board or on shore. Introducing autonomous operations, independently of the degree, implies changes to the existing roles of maritime personnel. The IMO distinction includes automation, remoting and autonomy under a joint term autonomy. In this paper, we separate automation from intelligent technologies that are considered disruptive for the maritime industry. This separation is useful from Human Factors perspective because humanautonomy interactions are expected to be different than those between human and automation. Automation is an "execution of machine function that was previously carried out by a human"(Parasuraman & Riley, 1997, p. 231) . To refer to autonomy, we use the umbrella term Robotic, Intelligent, Autonomous (RIA) technology, coined by the International Organization for Standardization (ISO) . ISO uses the term RIA to describe different forms of technology with an ability to self-regulate and self-govern (ISO, 2021) . These characteristics allow bilateral, context-dependent and, in result, social communication between RIA and humans. While automation is a tool for replacing human work, RIA can be viewed as something more – even as an assistant or a teammate. Therefore, trust appears to be even more relevant aspect of interaction between humans and technology. Introducing RIA to a sociotechnical system increases complexity

FAQs

Why is trust a key issue for autonomous maritime operations?

Trust is crucial because RIA technologies enable bilateral, context-dependent communication with humans, making interaction more social and requiring humans to rely on system judgments. Increased complexity also adds more interfaces and information that must be handled responsibly.

How does the paper define and distinguish autonomy-related technologies?

The paper separates automation from intelligent technologies that are considered disruptive for maritime use. It uses the ISO umbrella term Robotic, Intelligent, Autonomous (RIA) technology to cover systems capable of self-regulation and self-governance.

What trust-issue categories did the HUMANE workshops identify, and what do they suggest for training?

The workshops identified four meta-categories: operational trust, trustworthiness, organisational trust, and social acceptance. The findings point to improvements such as measuring perceived autonomy, using explainable AI, addressing linguistic and cultural communication aspects, and facilitating calibration of trust.

Aug 26, 2026

Uploaded by Oliver Hayes