Robots have become increasingly present in our daily lives, and we see several settings wherein this technology might positively impact people. One of these settings, and the focus of this paper, is the museum. This paper provides an overview of the literature on how robots can help people within the museum context and explores the state-of-the-art methods or technologies employed by researchers to allow robots to interact within this environment. Five key features were found to be essential to integrate into social museum robots: social navigation (including safe and successful motions); perception (how robots can use visual skills to understand the environment); speech (verbal communication with visitors); gestures (non-verbal cues) and behavior generation (how the robot combines the four skills to adopt different behaviors). The paper subsequently analyzes the integration of these features into robotic systems for a complete and appropriate collection of behaviors in the museum setting.

Personalization and localization are important when developing social robots for different sectors, including education, industry, healthcare or restaurants. This allows for an adjustment of robot behaviors according to the needs, preferences or personality of an individual when referring to personalization or to the social conventions or the culture of a country when referring to localization. However, there are different models that enable personalization and localization presented in the current literature, each with their advantages and drawbacks. This work aims to help researchers in the field of social robotics by reviewing and analyzing different papers in this domain. We specifically focus our review by exploring different robots that employ distinct models for the adaptation of the robot to its environment. Additionally, we study an array of methods used to adapt the nonverbal and verbal skills of social robots, including state-of-the-art techniques in artificial intelligence.

Background: Sentiment expression and detection are crucial for effective and empathetic human-robot interaction. Previous work in this field often focuses on non-verbal emotion expression, such as facial expressions and gestures. Less is known about which specific prosodic speech elements are required in human-robot interaction. Our research question was: what prosodic elements are related to emotional speech in human-computer/robot interaction?Methods: The scoping review was conducted in alignment with the Arksey and O’Malley methods. Literature was identified from the SCOPUS, IEEE Xplore, ACM Digital Library and PsycINFO databases in May 2021. After screening and de-duplication, data were extracted into an Excel coding sheet and summarised.Results: Thirteen papers, published from 2012 to 2020 were included in the review. The most commonly used prosodic elements were tone/pitch (n = 8), loudness/volume (n = 6) speech speed (n = 4) and pauses (n = 3). Non-linguistic vocalisations (n = 1) were less frequently used. The prosodic elements were generally effective in helping to convey or detect emotion, but were less effective for negative sentiment (e.g., anger, fear, frustration, sadness and disgust).Discussion: Future research should explore the effectiveness of commonly used prosodic elements (tone, loudness, speed and pauses) in emotional speech, using larger sample sizes and real-life interaction scenarios. The success of prosody in conveying negative sentiment to humans may be improved with additional non-verbal cues (e.g., coloured light or motion). More research is needed to determine how these may be combined with prosody and which combination is most effective in human-robot affective interaction.

Providing a robot with the capabilities of understanding and effectively adapting its behaviour based on human mental states is a critical challenge in Human–Robot Interaction, since it can significantly improve the quality of interaction between humans and robots. In this work, we investigate whether considering human mental states in the decision-making process of a robot improves the transparency of its behaviours and mitigates potential human’s false beliefs about the environment during collaborative scenarios. We used Bayesian inference within a Hierarchical Reinforcement Learning algorithm to include human desires and beliefs into the decision-making processes of the robot, and to monitor the robot’s decisions. This approach, which we refer to as Hierarchical Bayesian Theory of Mind, represents an upgraded version of the initial Bayesian Theory of Mind, a probabilistic model capable of reasoning about a rational agent’s actions. The model enabled us to track the mental states of a human observer, even when the observer held false beliefs, thereby benefiting the collaboration in a multi-goal task and the interaction with the robot. In addition to a qualitative evaluation, we conducted a between-subjects study (110 participants) to evaluate the robot’s perceived Theory of Mind and its effects on transparency and false beliefs in different settings. Results indicate that a robot which considers human desires and beliefs increases its transparency and reduces misunderstandings. These findings show the importance of endowing Theory of Mind capabilities in robots and demonstrate how these skills can enhance their behaviours, particularly in human–robot collaboration, paving the way for more effective robotic applications.