International Journal of Social Robotics最新文献

Increasingly often robots are deployed in human environments, where they will encounter people. An example of a challenge robots encounter is crossing paths with a human. Based on human-robot proxemics research one would expect that people would keep a certain distance to maintain an appropriate comfort level. However it is unclear whether this also holds for crossing scenarios between a robot and a person. In the first experiment presented in this paper, a humanoid robot crossed paths with a person in which the crossing angle and acceleration of the robot were manipulated. Results showed that participants deviated more from a straight path when the robot arrived earlier at the crossing point compared to the other trials and when it accelerated or when the robot itself deviated from a straight path. If participants had to deviate from their path, it was regarded as less comfortable and it required more effort. In the second experiment, an autonomous guided vehicle was used, and we tested the moving speed of the robot. Similar to the first experiment, when the robot kept a straight path or stopped, it was regarded as the most comfortable. The results show that it is more comfortable if a robot does not change its direction while crossing paths with the robot. These findings indicate that perceived comfort is not merely determined by distance, but is more strongly affected by how predictable the robot is.

Co-speech gestures have significant impacts on conveying information. For social agents, producing realistic and smooth gestures are crucial to enable natural interactions with humans, which is a challenging task depending on many impact factors (e.g., speech audio, content, and the interacting person). In this paper, we tackle the cross-modal fusion problem through a novel fusion mechanism for end-to-end learning-based co-speech gesture generation. In particular, we facilitate parallel directional cross-modal transformers, and an interactive and cascaded 2D attention module, to achieve selective fusion of the gesture-related cues. Besides, we propose new metrics to evaluate gesture diversity and speech-gesture correspondence, without 3D pose annotation requirements. Experiments on a public dataset indicate that the proposed method can successfully produce diverse human-like poses, which outperform the other competitive state-of-the-art methods, with the evaluations conducted both objectively and subjectively.

This paper contributes to the understanding of child-robot interaction through the investigation of child interactions with and anthropomorphization of humanoid robots when manipulating robot-related variables such as behavior and gender. In this study, children observe a robot demonstration in a classroom setting, during which the robot showcases either assertive or submissive behavior and is attributed a gender, either robot-female or robot-male. Afterwards, participant anthropomorphization is measured using the Attributed Mental States Questionnaire (AMS-Q). Results suggest that when prompted to select a response directed at the robot, children used significantly more commanding phrases when addressing the assertively behaving robot when compared to the submissively behaving robot. Further, younger children ages 7–9 anthropomorphize robots at a higher degree than older children 10–12 and assertive behavior from the robot lead to higher rates of anthropomorphization. Results also suggest that children are more likely to respond to female robots in an imperative way than male robots. This widened understanding of child perception of and interaction with humanoid robots can contribute to the design of acceptable robot interaction patterns in various settings.

As social robots are projected to become an integral part of human life in the coming decades, their ability to adapt movement and trajectory when in proximity to people is essential for ensuring social acceptance during human-robot interaction. A key aspect of this adaptability involves predicting and anticipating human intents during robot navigation. Despite significant strides in the social navigation of autonomous robots within human environments, opportunities for advancements in related algorithms persist. This paper presents a novel real-time path trajectory optimization algorithm for socially aware robot navigation, grounded in the social elastic band concept, incorporating prediction and anticipation of human movements to adapt its forward velocity. Building upon the elastic band framework introduced in the 1990s for adapting robot trajectories in dynamic environments, our proposal of social elastic band differentiates between objects and human presence. This distinction allows for the definition of social interaction spaces and their relationship to the elastic band, facilitating the generation of socially accepted paths that rapidly adapt to environmental changes without causing a disturbance. Integrated into the SNAPE social navigation framework, the algorithm has been tested and validated through simulations and real-world experiments in various environments.

Communication among some older adults is affected by cognitive and mobility impairments. This increases isolation, particularly for those residing in care homes, and leads to accelerated cognitive decline. Previous research has leveraged assistive robots to promote recreational routines and communication among older adults, with the robot leading the interaction. However, older adults could have more agency in the interaction, as robots could extend elders’ intentions and needs. Therefore, we explored an approach whereby the robot’s agency is shifted to the older adults who lead the interaction by commanding a robot’s actions using interactive physical blocks (tangible blocks). We conducted sessions with 22 care home dwellers where they could exchange messages and objects using the robot. Based on older adults’ observed behaviors during the sessions and perspectives gathered from interviews with geriatric professionals, we reflect on the opportunities and challenges for increased user agency and the asymmetries that emerged from differing abilities and personality traits. Our qualitative results highlight the potential of robotic approaches to extend the agency and communication of older adults, anchored on human values, such as the exchange of affection, collaboration, and competition.

Examining personality traits can enhance the likelihood of a successful interaction between humans and robots in forthcoming work settings. Employing the emic/etic approach stands out as a crucial method for investigating personality types in the context of future environments. Currently, no study has explored the impact of this approach on individuals’ willingness to engage with a robot. In the present study, our aim is to determine whether emic characteristics can influence the connection between etic traits and the willingness to use a robot. In the current study, 367 male workers participated. All data were collected using valid and reliable questionnaires. The Five-Factor model of personality was regarded as etic personality characteristics, while the moderating roles of technology affinity and STARA were assessed as emic personality characteristics. The analytical process followed the method presented by Hayes et al. for analyzing moderators. Technology affinity, as a primary emic factor, exerts a moderating influence on the association between neuroticism, openness, agreeableness, conscientiousness, and the willingness to use robots. Conversely, STARA serves as a mediator exclusively in the relationship with neuroticism among workers. Notably, extroversion does not exhibit mediation with any of the emic factors. Both emic and etic personality characteristics were recognized as significant facilitators of the inclination to use robots. In addition to technology affinity and STARA, it is advisable to explore new emic traits and their interactive effects with etic personality characteristics.

This study explores how much current mainstream Robot-Assisted Language Learning (RALL) systems produce outcomes compared to human tutors instructing a typical English conversation lesson. To this end, an experiment was conducted with 26 participants divided in RALL (14 participants) and human tutor (12 participants) groups. All participants took a pre-test on the first day, followed by 30 min of study per day for 7 days, and 3 post-tests on the last day. The test results indicated that the RALL group considerably improved lexical/grammatical error rates and fluency of speech compared to that for the human tutor group. The other characteristics, such as rhythm, pronunciation, complexity, and task achievement of speech did not indicate any differences between the groups. The results suggested that exercises with the RALL system enabled participants to commit the learned expressions to memory, whereas those with human tutors emphasized on communication with the participants. This study demonstrated the benefits of using RALL systems that can work well in lessons that human tutors find hard to teach.

In recent years, applications of social robots as the operator’s avatar have been widely studied for remote conversation with rich nonverbal information. Having another side-participant robot beside the avatar robot of the operator was found to be effective for providing long-lasting backchannels to the interlocutor. The side-participant robot is also expected to play a role in assisting human participation in multiparty conversations. However, such a focus has not been applied to remote conversations with multiple robots. Here, we propose a multiple-robot telecommunication system with which the operator can use a side-participant robot to assist conversation that is developed by the operator through the main speaker robot to verify its effectiveness. In the laboratory experiment where the subjects were made to feel stressed by being forced to provide rude questions to the interlocutor, the proposed system was shown to reduce guilt and to improve the overall mood of operators. The result encourages the application of a multi robot remote conversation system to allow the user to participate in remote conversations with less anxiety of potential failure in maintaining the conversation.

As robots become more common, people interact with them individually, with strangers, and with friends. For example, when coming across a robot in a mall, a family might ask it for instructions. An individual person might hesitate to interact with the robot until they see another person interacting, and then explore the robot together. Although human–robot interaction (HRI) research has recently uncovered the importance of examining differences in group behavior toward robots versus individuals’ behavior, thus far, most HRI research has not distinguished behavior based on group type (e.g., stranger, companion). In this online lab-based study, we explore how individuals, strangers, and companions collaborate with robot teammates. We test competing hypotheses: (1) More cohesive companion groups will form a human subgroup and exclude the robots more than strangers or individuals, vs. (2) More cohesive companion groups will provide social support to interact better with the novel robotic technology than strangers or individuals. In this cooperative context in which participants were required to interact with the robot, results supported H1: the subgroup hypothesis. Based on these findings, people deploying robots should note that if people are required to interact with the robots, the interactions may not go as smoothly for companion groups compared to stranger groups or individuals.

This study examined the effects of explanation strategy (global explanation vs. deductive explanation vs. contrastive explanation) and autonomy level (high vs. low) of explainable agents on human–AI collaborative decision-making. A 3 × 2 mixed-design experiment was conducted. The decision-making task was a modified Mahjong game. Forty-eight participants were divided into three groups, each collaborating with an agent with a different explanation strategy. Each agent had two autonomy levels. The results indicated that global explanation incurred the lowest mental workload and highest understandability. Contrastive explanation required the highest mental workload but incurred the highest perceived competence, affect-based trust, and social presence. Deductive explanation was found to be the worst in terms of social presence. The high-autonomy agents incurred lower mental workload and interaction fluency but higher faith and social presence than the low-autonomy agents. The findings of this study can help practitioners in designing user-centered explainable decision-support agents and choosing appropriate explanation strategies for different situations.