Child-robot Interaction Research Articles

The JIBO Kids Corpus: A speech dataset of child-robot interactions in a classroom environment.

This paper describes an original dataset of children's speech, collected through the use of JIBO, a social robot. The dataset encompasses recordings from 110 children, aged 4-7 years old, who participated in a letter and digit identification task and extended oral discourse tasks requiring explanation skills, totaling 21 h of session data. Spanning a 2-year collection period, this dataset contains a longitudinal component with a subset of participants returning for repeat recordings. The dataset, with session recordings and transcriptions, is publicly available, providing researchers with a valuable resource to advance investigations into child language development.

Preliminary study on the feasibility of approximating children's engagement level from their emotions estimation by a picture-based, three-model AI in a family-robot cohabitation scenario

Haru4Kids (H4K) is a system that emulates the physical, social, family-oriented robot Haru, designed with the goal of cohabitating with children in their homes for extended periods of time. In a previous experiment [Garcia GA, Perez G, Levinson L, et al. Living with Haru4Kids: study on children's activity and engagement in a family-robot cohabitation scenario. In: 2023 IEEE ROMAN; 2023 Aug. p. 1428–1435], seven families kept H4K for a span of two weeks in their homes. Throughout this period of cohabitation, we collected child-robot interaction data, including images that were later hand-annotated to estimate user engagement. In this present work, we used a novel AI-based, four-stage framework available from Roboflow for the automatic estimation of children's level of engagement from their inferred emotions. We did a deep study of the performance and behaviour of that framework over our dataset of users' pictures and characterized its response in order to understand its advantages and limitations, including the technique used to translate emotions into engagement levels. We also tested a different approach for that mapping, using a machine learning technique based on Support Vector Machine (SVM). The framework yielded promising results just ‘off-the-shelf’: 0.47–0.68 accuracy and 0.46–0.70 F1 using the original mapping, and 0.39–0.75 and 0.37–0.78 respectively using SVM. Therefore, we propose this emotion-based approach for engagement estimation from pictures.

A multi-site language study on child-robot dialogues

Child-robot interaction (CRI) has been mostly studied in labs and classroom settings. In this work, we share a CRI language processing study carried out in children's homes, analyzing how each language processing layer performs with children at home with no researcher present. We carried out an experiment in two phases: a first iteration with 7 families in Spain cohabiting with a simulated robot for 2 weeks and a second iteration with 31 families: 11 in Spain, 9 in Ukraine and 11 in the USA completing 4 cohabitation weeks at each country. The total number of children involved in the experiments was 55, ranging from 5 to 13 years old. Our goal in this study is to evaluate the performance of voice recognition, language understanding and dialogue management when children interact with a robot at home. Our results indicate that dialogue management capabilities are becoming the key element in the language processing pipeline; they also denote that the dialogue engine should include mixed-initiative capabilities and show the relative usage of different common built-in intents.

Intelligent Agents at School—Child–Robot Interactions as an Educational Path

The pervasiveness of technologies leads us to talk about a code society. From an educational point of view, coding, computational thinking, and educational robotics are an open possibility. Nevertheless, new elements such as artificial intelligence are rapidly changing educational technology perspectives. In this work, we will analyze school policies and theoretical bases in order to understand if, and under what kind of, condition coding, computational thinking, and educational robotics still represent the qualifying elements of a framework for digital literacy and digital citizenship.

Balancing Human Likeness in Social Robots: Impact on Children’s Lexical Alignment and Self-disclosure for Trust Assessment

While there is evidence that human-like characteristics in robots could benefit child-robot interaction in many ways, open questions remain about the appropriate degree of human likeness that should be implemented in robots to avoid adverse effects on acceptance and trust. This study investigates how human likeness, appearance and behavior, influence children’s social and competency trust in a robot. We first designed two versions of the Furhat robot with visual and auditory human-like and machine-like cues validated in two online studies. Secondly, we created verbal behaviors where human likeness was manipulated as responsiveness regarding the robot’s lexical matching. Then, 52 children (7-10 years old) played a storytelling game in a between-subjects experimental design. Results show that the conditions did not affect subjective trust measures. However, objective measures showed that human likeness affects trust differently. While low human-like appearance enhanced social trust, high human-like behavior improved children’s acceptance of the robot’s task-related suggestions. This work provides empirical evidence on manipulating facial features and behavior to control human likeness in a robot with a highly human-like morphology. We discuss the implications and importance of balancing human likeness in robot design and its impacts on task performance, as it directly impacts trust-building with children.

“No, I Won't Do That.” Assertive Behavior of Robots and its Perception by Children

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.

Lessons in Developing a Behavioral Coding Protocol to Analyze In-the-Wild Child–Robot Interaction Events and Experiments

Behavioral analyses of in-the-wild HRI studies generally rely on interviews or visual information from videos. This can be very limiting in settings where video recordings are not allowed or limited. We designed and tested a vocalization-based protocol to analyze in-the-wild child–robot interactions based upon a behavioral coding scheme utilized in wildlife biology, specifically in studies of wild dolphin populations. The audio of a video or audio recording is converted into a transcript, which is then analyzed using a behavioral coding protocol consisting of 5–6 categories (one indicating non-robot-related behavior, and 4–5 categories of robot-related behavior). Refining the code categories and training coders resulted in increased agreement between coders, but only to a level of moderate reliability, leading to our recommendation that it be used with three coders to assess where there is majority consensus, and thereby correct for subjectivity. We discuss lessons learned in the design and implementation of this protocol and the potential for future child–robot experiments analyzed through vocalization behavior. We also perform a few observational behavior analyses from vocalizations alone to demonstrate the potential of this field.

‘How Would you Score Yourself?’: The Effect of Self-assessment Strategy Through Robots on Children’s Motivation and Performance in Piano Practice

This research examines how to design social robots to support self-regulated learning skills for piano practice. More specifically, a social robot is used to provide feedback to children and initiate self-assessment. To assess the impact of this approach on children’s motivation and performance, we conducted an experiment in a music school where 50 children practiced with both a self-assessment and a non-evaluative robot. Results showed that when the children interacted with the self-assessment robot they had higher motivation and better performance than when they interacted with the non-evaluative robot. Furthermore, interaction effects were found between the robot conditions, the children’s learning stages, and their gender regarding their motivation and rhythm performance. Overall, the study demonstrates a positive influence of robot-initiated self-assessment on children’s musical instrument practice and provided insights for personalized child-robot interaction design.

Beyond the Session: Centering Teleoperators in Socially Assistive Robot-Child Interactions Reveals the Bigger Picture

Socially assistive robots play an effective role in children's therapy and education. Robots engage children and provide interaction that is free of the potential judgment of human peers and adults. Research in socially assistive robots for children generally focuses on therapeutic and educational outcomes for those children, informed by a vision of autonomous robots. This perspective ignores therapists and educators, who operate these robots in practice. Through nine interviews with individuals who have used robots to deliver socially assistive services to neurodivergent children, we (1) define a dual-cycle model of therapy that helps capture the domain expert view of therapy, (2) identify six core themes of teleoperator needs and patterns across these themes, (3) provide high-level guidelines and detailed recommendations for designing teleoperated socially assistive robot systems, and (4) outline a vision of robot-assisted therapy informed by these guidelines and recommendations that centers teleoperators of socially assistive robots in practice.

Child-Robot Interaction in Healthcare: Opportunities, Challenges, and Resolutions

Robots are increasingly permeating the healthcare system, but limited studies exist in the field of child-robot interaction in healthcare.

Can You Dance? A Study of Child–Robot Interaction and Emotional Response Using the NAO Robot

This retrospective study presents and summarizes our long-term efforts in the popularization of robotics, engineering, and artificial intelligence (STEM) using the NAO humanoid robot. By a conservative estimate, over a span of 8 years, we engaged at least a couple of thousand participants: approximately 70% were preschool children, 15% were elementary school students, and 15% were teenagers and adults. We describe several robot applications that were developed specifically for this task and assess their qualitative performance outside a controlled research setting, catering to various demographics, including those with special needs (ASD, ADHD). Five groups of applications are presented: (1) motor development activities and games, (2) children’s games, (3) theatrical performances, (4) artificial intelligence applications, and (5) data harvesting applications. Different cases of human–robot interactions are considered and evaluated according to our experience, and we discuss their weak points and potential improvements. We examine the response of the audience when confronted with a humanoid robot featuring intelligent behavior, such as conversational intelligence and emotion recognition. We consider the importance of the robot’s physical appearance, the emotional dynamics of human–robot engagement across age groups, the relevance of non-verbal cues, and analyze drawings crafted by preschool children both before and after their interaction with the NAO robot.

Measuring Engagement in Robot-Assisted Therapy for Autistic Children.

Children with autism face a range of challenges when it comes to verbal and nonverbal communication. It is essential that children participate in a variety of social, educational, and therapeutic activities to acquire knowledge that is essential for cognitive and social development. Recent studies have shown that children with autism may be interested in playing with an interactive robot. The robot can engage these children in ways that demonstrate and train essential aspects of human interaction, guiding them in therapeutic sessions to practice more complex forms of interaction found in social human-to-human interactions. This study sets out to investigate Robot-Assisted Autism Therapy (RAAT) and the use of artificial intelligence (AI) approaches for measuring the engagement of children during therapy sessions. The study population consisted of five native Arabic-speaking autistic children aged between 4 and 11 years old. The child-robot interaction was recorded by the robot camera and later used for analysis to detect engagement. The results show that the proposed system offers some accuracy in measuring the engagement of children with ASD. Our findings revealed that robot-assisted therapy is a promising field of application for intelligent social robots, especially to support autistic children in achieving their therapeutic and educational objectives.

Promoting Honesty in Children, or Fostering Pathological Behaviour?

Promoting Honesty in Children, or Fostering Pathological Behaviour?

It Takes Two: Using Co-creation to Facilitate Child-Robot Co-regulation

While interacting with a social robot, children have a need to express themselves and have their expressions acknowledged by the robot—a need that is often unaddressed by the robot, due to its limitations in understanding the expressions of children. To keep the child-robot interaction manageable, the robot takes control, undermining children’s ability to co-regulate the interaction. Co-regulation is important for having a fulfilling social interaction. We developed a co-creation activity that aims to facilitate more co-regulation. Children are enabled to create sound effects, gestures, and light animations for the robot to use during their conversation. A crucial additional feature is that children are able to coordinate their involvement of the co-creation process. Results from a user study (n= 59 school children, 7–11 years old) showed that the co-creation activity successfully facilitated co-regulation by improving children’s agency. It also positively affected the acceptance of the robot. We furthermore identified five distinct profiles detailing the different needs and motivations children have for the level of involvement they chose during the co-creation process.

Dysgraphia detection based on convolutional neural networks and child-robot interaction

Dysgraphia is a disorder of expression with the writing of letters, words, and numbers. Dysgraphia is one of the learning disabilities attributed to the educational sector, which has a strong impact on the academic, motor, and emotional aspects of the individual. The purpose of this study is to identify dysgraphia in children by creating an engaging robot-mediated activity, to collect a new dataset of Latin digits written exclusively by children aged 6 to 12 years. An interactive scenario that explains and demonstrates the steps involved in handwriting digits is created using the verbal and non-verbal behaviors of the social humanoid robot Nao. Therefore, we have collected a dataset that contains 11,347 characters written by 174 participants with and without dysgraphia. And through the advent of deep learning technologies and their success in various fields, we have developed an approach based on these methods. The proposed approach was tested on the generated database. We performed a classification with a convolutional neural network (CNN) to identify dysgraphia in children. The results show that the performance of our model is promising, reaching an accuracy of 91%.

Children’s Digital and Non-Digital Play Practices with Cozmo, the Toy Robot

Children’s Digital and Non-Digital Play Practices with Cozmo, the Toy Robot

The Effects of Social Presence and Familiarity on Children-Robot Interactions.

In children-robot interactions, an impression of a robot's "social presence" (i.e., an interactive agent that feels like a person) links positively to an improved relationship with the robot. However, building relationships takes many exposures, and there is an intellectual gap in how social presence and familiarity collaborate in modulating children-robot relationships. We investigated whether social presence altered over time, how repeated exposure and social presence affected rapport, and how social presence would modulate children's attitudes toward the robot. Fourteen children (four female, age = 10.79 ± 1.12) interacted with a companion robot for four days in spontaneous interactions. The findings revealed that children who perceived the robot as having a higher social presence developed a stronger rapport than those who perceived a lower social presence. However, repeated encounters did not change the children's perceptions of the robot's social presence. Children rated higher rapport after repeated interactions regardless of social presence levels. This suggests that while a higher social presence initially elevated the positive relationship between children and the robot, it was the repeated interactions that continued solidifying the rapport. Additionally, children who perceived a higher social presence from the robot felt less relational uneasiness about their relationship with robots. These findings highlight the importance of robots' social presence and familiarity in promoting positive relationships in children-robot interaction.

Child–robot interaction: designing robot mediation to facilitate friendship behaviors

ABSTRACT This research explored if a social robot would play a role in facilitating the development of friendship between young children while they engage in playful learning. Grounded in child–robot interaction and child development literature, we instantiated four sessions of triadic interaction activities among two children and a robot, where the robot mediated the children’s collaborative interactions. There were two types of robot-mediated activities (Conversational and Tablet assisted) with each type having two sessions and each session taking approx. 20 minutes, depending on the children. The activities were deployed with ten children (aged five to six) in an after-school program in a rural public school in the U.S. twice a week for two weeks. We video-recorded the sessions and later annotated these recordings for analysis. The friendship development between the children in a pair was observed in terms of five behavioral categories (liking, togetherness, parity, agreement, and co-construction). The results showed that both conversational and tablet-assisted robot mediation contributed complementarily to friendship development among the children.

An Analysis of Research Trends in the Use of Social Robots in Early Childhood Institutions

이 논문은 소셜 로봇의 보급화에 주목하여 영유아 교육기관에서의 소셜 로봇활용 연구 경향을 분석하는 데 목적이 있다. 이에 2008년부터 2022년까지 국내 학술지에 게재된 논문 38편의 게재 경향, 연구 특성과 로봇활용 특성을 분석하였다. 분석결과, 발행연도별 연구편수는 2011 ~ 2013년에 그 수가 가장 높고 이후 감소하다 최근 들어 다시 증가하는 경향을 보였다. 연구주제 및 연구대상에서는 영유아 대상의 프로그램 효과성 검증과 아동-로봇 상호작용 연구가 가장 많고, 그다음으로 교사 대상의 로봇활용 태도 및 인식 연구 순으로 나타났다. 연구주제에 따른 특정 연구방법이 편중되어 있었으며, 연구에 사용된 로봇명은 제니보와 아이로비큐가 대다수를 차지하였다. 로봇활용 기간 범위는 한 달 미만에서 다섯 달 이상이었으며, 대체로 로봇 투입 기간이 길어짐에 따라 로봇활용 빈도가 낮아졌다. 이를 바탕으로 최근 영유아 기관에서의 소셜 로봇활용 연구가 증가하고 있지만 타학문과의 협력을 통해 연구가 활발히 이루어질 필요가 있으며, 연구주제와 연구대상, 연구방법론적 측면을 보완하고, 로봇활용 특성에 따른 로봇활용 프로그램 효과성 검증의 필요성을 제기하였다.The purpose of this study is to analyze research trends in the use of social robots in early childhood institutions. A total of 38 research publications published in South Korean journals from 2008 to 2022 were analyzed in the criteria of publication trends and features of research. As a result, the number of publications was the highest in 2011 to 2013, decreased thereafter, and recently showed a tendency to increase again. In terms of research topics and participants, child-directed programs and child-robot interaction studies were the most common, followed by teacher’s attitudes and perceptions of robot use. Research methods were biased according to the research subject, and Genibo and IrobiQ were the most used robots. The period of robot use ranged widely and the frequency of robot use generally decreased as the period of robot use became longer. Based on these findings, although research on the use of social robots in early childhood institutions has recently increased, research needs to be conducted actively through cooperation with other disciplines and to be supplemented in the aspects of research topics, participants, and methodologies.

Correction to: Ethical Considerations from Child-Robot Interactions in Under-Resourced Communities

Correction to: Ethical Considerations from Child-Robot Interactions in Under-Resourced Communities