Extending Three Existing Models to Analysis of Trust in Automation: Signal Detection, Statistical Parameter Estimation, and Model-Based Control

Trust as Extended Control: Human-Machine Interactions as Active Inference

In order to interact seamlessly with robots, users must infer the causes of a robot's behavior-and be confident about that inference (and its predictions). Hence, trust is a necessary condition for human-robot collaboration (HRC). However, and despite its crucial role, it is still largely unknown how trust emerges, develops, and supports human relationship to technological systems. In the following paper we review the literature on trust, human-robot interaction, HRC, and human interaction at large. Early models of trust suggest that it is a trade-off between benevolence and competence; while studies of human to human interaction emphasize the role of shared behavior and mutual knowledge in the gradual building of trust. We go on to introduce a model of trust as an agent' best explanation for reliable sensory exchange with an extended motor plant or partner. This model is based on the cognitive neuroscience of active inference and suggests that, in the context of HRC, trust can be casted in terms of virtual control over an artificial agent. Interactive feedback is a necessary condition to the extension of the trustor's perception-action cycle. This model has important implications for understanding human-robot interaction and collaboration-as it allows the traditional determinants of human trust, such as the benevolence and competence attributed to the trustee, to be defined in terms of hierarchical active inference, while vulnerability can be described in terms of information exchange and empowerment. Furthermore, this model emphasizes the role of user feedback during HRC and suggests that boredom and surprise may be used in personalized interactions as markers for under and over-reliance on the system. The description of trust as a sense of virtual control offers a crucial step toward grounding human factors in cognitive neuroscience and improving the design of human-centered technology. Furthermore, we examine the role of shared behavior in the genesis of trust, especially in the context of dyadic collaboration, suggesting important consequences for the acceptability and design of human-robot collaborative systems.

Supporting Drivers of Partially Automated Cars through an Adaptive Digital In-Car Tutor

Drivers struggle to understand how, and when, to safely use their cars’ complex automated functions. Training is necessary but costly and time consuming. A Digital In-Car Tutor (DIT) is proposed to support drivers in learning about, and trying out, their car automation during regular drives. During this driving simulator study, we investigated the effects of a DIT prototype on appropriate automation use and take-over quality. The study had three sessions, each containing multiple driving scenarios. Participants needed to use the automation when they thought that it was safe, and turn it off if it was not. The control group read an information brochure before driving, while the experiment group received the DIT during the first driving session. DIT users showed more correct automation use and a better take-over quality during the first driving session. The DIT especially reduced inappropriate reliance behaviour throughout all sessions. Users of the DIT did show some under-trust during the last driving session. Overall, the concept of a DIT shows potential as a low-cost and time-saving solution for safe guided learning in partially automated cars.

Positive bias in the ‘Trust in Automated Systems Survey’? An examination of the Jian et al. (2000) scale

Measuring trust in technology is a mainstay in Human Factors research. While trust may not perfectly predict reliance on technology or compliance with alarm signals, it is routinely used as a design consideration and assessment goalpost. Several methods of measuring trust have been employed in the past decades, but one self-report measure stands out due to its popular use, the Trust in Automated Systems Survey (Jian, Bisantz, & Drury, 2000). We conducted a study to assess whether the survey could create biased responses, and found evidence the original scale is in fact skewed toward positive ratings. Assessing the literature revealed the survey has been used in unaltered form across at least 100 different reports and remains frequently administered – therefore, the potential impact of this bias may be widespread. Future directions, considerations, and caveats for our assessment, and for using this scale, are discussed.

Individual Differences in Attributes of Trust in Automation: Measurement and Application to System Design

Computer-based automation of sensing, analysis, memory, decision-making, and control in industrial, business, medical, scientific, and military applications is becoming increasingly sophisticated, employing various techniques of artificial intelligence for learning, pattern recognition, and computation. Research has shown that proper use of automation is highly dependent on operator trust. As a result the topic of trust has become an active subject of research and discussion in the applied disciplines of human factors and human-systems integration. While various papers have pointed to the many factors that influence trust, there currently exists no consensual definition of trust. This paper reviews previous studies of trust in automation with emphasis on its meaning and factors determining subjective assessment of trust and automation trustworthiness (which sometimes but not always are regarded as an objectively measurable properties of the automation). The paper asserts that certain attributes normally associated with human morality can usefully be applied to computer-based automation as it becomes more intelligent and more responsive to its human user. The paper goes on to suggest that the automation, based on its own experience with the user, can develop reciprocal attributes that characterize its own trust of the user and adapt accordingly. This situation can be modeled as a formal game where each of the automation user and the automation (computer) engage one another according to a payoff matrix of utilities (benefits and costs). While this is a concept paper lacking empirical data, it offers hypotheses by which future researchers can test for individual differences in the detailed attributes of trust in automation, and determine criteria for adjusting automation design to best accommodate these user differences.