Theories of embodied learning emphasize that learning processes are grounded in bodily actions and interactions with the environment, suggesting that movements play a fundamental role in problem solving, decision making, and learning. This perspective holds particular relevance for making-based learning settings, where patterns of movement and spatial engagement can reveal strategic expertise. Prior research has examined distinctions between students who learned and did not learn, but manual coding of actions presents scalability and real-time application challenges. To address this gap, we develop a computer vision–based analysis pipeline for automated detection and characterization of hand movements during complex assembly tasks. In an exploratory study, we apply this approach to video data of students engaged in the assembly of a differential gearbox, quantifying metrics such as amount and speed of movement. Results indicate that learners show fewer right-hand movements than novices and exhibit reduced movement speed, with a progressive decline in speed as the task unfolds. Non-learners, by contrast, display more uneven hand movement speed. These findings, while preliminary, highlight measurable differences in actions of learners and non-learners, and therefore have potential implications for learning support. Specifically, the ability to computationally distinguish movement profiles can inform the design of adaptive learning interventions, providing real-time performance assessment and targeted feedback for making-based learning.

Embedding social robots in educational and/or childcare settings has potential to engage children while supporting caregivers. However, little is known about the practical, long-term integration of robots in such settings. Our work addresses this gap through an ethnographic study that adopts a year-long perspective toward a socially assistive robot in an inclusive daycare with a four-month robot deployment phase. Through Thematic Analysis, we highlight children's ability to develop a variety of self-determined interactions with the robot, while unveiling caregivers' need for its integration into structured routines such as meal times, as well as challenges for robustness of the robot in our research setting. On this basis, we make recommendations for the design of social robots for daycare environments.

When witnessing the abuse of others, humans generally exhibit emotional responses. A large number of studies in Human-Robot-Interaction (HRI) have shown that humans also react with sympathy when robots are abused, but most of these insights come from controlled laboratory studies using short videos and student samples. To complement existing research with observations drawn from real-world online discussions, this paper presents a sentiment analysis of 103,413 YouTube comments on videos depicting abuse of animal-like, humanoid, and cart-shaped robots. To validate our sentiment classification, we analysed the comments using a lexicon-based tool, two fine-tuned language models, and three general-purpose state-of-the-art large language models (LLMs). The comparison yielded interesting results: LLMs generally classified science-fiction–related comments, e.g., references to dystopian TV shows, as negative, while lexicon and fine-tuned models mainly labelled them as neutral. The six models agreed on the classification of a total of 27,427 comments, which we used to explore the sentiment expressions occurring across videos featuring robots with different physical forms. Our findings provide large-scale, ecologically valid insights into how emotional responses to robot abuse are expressed and analysed in online video platforms.

Sonic Human-Robot Interaction aims at equipping robots with the ability to convey emotions and intentions via sounds. Such sounds are typically handcrafted by human experts, which results in expensive, small sound sets with limited expressivity. To overcome these limitations, in this paper we propose (i) a module for the analysis of the emotional content of robotic sounds (RER), which combines a Contrastive Learning model with a classifier and regressor, and (ii) an Evolutionary Strategy (ES) that enables the automated modulation of a sound to control for its emotional content. Experiments performed on 118 emotion-labelled sounds provided by a publicly available robot vocal library suggest that: (i) state-of-the-art solutions for emotion recognition in speech or musical samples cannot be directly applied to robotic sounds; (ii) our proposed RER module achieves an F1-score of 88 % in the categorical emotion classification of robotic sounds and correctly places sounds on the Valence-Arousal space; (iii) our proposed ES is a promising first step towards modulating a sound to convey a target emotion.

Large language models (LLMs), when integrated into social robots, have the potential to transform robot-assisted language learning by offering personalized, interactive communication. However, there is limited research exploring their potential to simultaneously reduce anxiety and enhance language-speaking skills among international university students, who often feel anxious when speaking a foreign language. This study addresses this gap by evaluating the impact of a humanoid robot powered by the OpenChat-3.5 LLM as a tandem partner for German language learning. Using a between-subjects design with 22 multilingual participants, two interaction conditions were tested: immersive (German-only) and bilingual (German-English). Our findings indicate that participants in the immersive mode reported experiencing significantly reduced perceived judgment by the robot compared to the bilingual mode. Although female participants showed a trend of greater improvement in learning gain, no significant gender differences were found. Open-ended feedback highlighted the need for enhanced contextual responses, slower speech rate, faster response times, and error corrections to enhance language speaking support. This study aims to advance social robots for learning by demonstrating the usage of generative AI in creating non-judgmental language practice scenarios.

Non-verbal children are often provided with non-verbal communication cards (NVCCs) to share their needs. However, in busy spaces that prioritize verbal communication, for example, a daycare, this may place them at a disadvantage in comparison with verbal peers. In this work, we introduce the concept of a Socially Mediating Robot (SMR), acting as a communicational intermediary that verbalizes NVCC input. Thereby, we want to explore whether non-verbal children can be empowered by giving them an additional way of expressing themselves in a dynamic, buzzing environment. In an exploratory interview study with six participants professionally working with children, we then provide first insights on the feasibility of our concept from the perspective of caregiving staff. Thereby, we lay the groundwork for evaluating the concept in an ongoing field study in an inclusive daycare context.

Persuasive social robots have the ability to influence human behaviour through social interaction, which makes them a valuable technological solution for all applications aiming to support a person's behaviour change. The recently proposed Child-Robot Relational Norm Intervention (CRNI) model intro-duces a new approach for persuasive social robotics, leveraging children's reluctance to disturb robots to promote behaviour change. Unlike traditional methods that rely on direct feedback or reminders, CRNI encourages children to self-monitor and self-correct improper behaviour, by making the robot express mild distress whenever the child exhibits the incorrect behaviour. This paper proposes the first implementation of the CRNI approach in a real HRI context and evaluates its effectiveness in improving children's handwriting posture. The evaluation includes two user studies: (i) a multi-session study with five children investigating the long-term impact of the approach, (ii) a controlled study with 29 children comparing CRNI to direct robot reminders. The results indicate that the CRNI model leads to more sustained posture correction compared to direct interventions. More broadly, our findings suggest that relational norm-based approaches can offer an effective yet less intrusive method for fostering positive behaviours in children.

Long-term child-robot interaction (CRI) in kindergartens is key for integrating robots into children’s daily lives and supporting their development. This unsupervised 4-weeks study explores how the robot Pepper can support phonological awareness (PA) in kindergarten children through the ‘Sound Game’, leveraging robot logs, questionnaires and open feedback by teachers and experts’ analysis of audio recordings. Our findings reveal a generally high engagement, with the game being played 12 times, for a total of 233 minutes. Teachers played a crucial role in mediating the game, especially for younger children, and noted its potential for long-term use. The game proved beneficial across age groups, highlighting its potential as a first step for sustained, effective PA development and the transformative impact of social robots in early childhood education.

This work explores whether canine-inspired motions on a quadruped robot effectively convey emotion and intent, specifically investigating (i) participants' recognition rate of the intended emotions & intents; (ii) the emotions that the robot movements elicit in the participants and (iii) the influence of prior experience with robots and dogs on the recognition rate. A user study involving 35 participants revealed that the movements designed to convey alert, neutral, and yes/agree exceeded the average human recognition rate for robotic emotional expressions through body gestures reported in prior work, while the analysis of the alignment between participants' emotional responses to the movements and their intended emotional content shed light on possible reasons for misinterpretations. Interestingly, prior experience with robots and dogs was found to have no significant impact on the recognition rates.

As social robots increasingly integrate into human society, their ability to sense our surroundings legitimately raises privacy concerns. The objective of this study is twofold. First, we explore the possibility of providing social robots with privacy-preserving sensing, i.e. the ability of extracting necessary sensory information while preserving users’ privacy. Second, we investigate whether the use of such privacy-preserving sensing as well as transparency with respect to the robot’s sensing capabilities can encourage individuals to self-disclose during human-robot conversations. A 2 × 2 between-subject experiment was conducted with 28 participants, who engaged in a conversation with the PixelBot robot. The results suggest that conversational robots can perform effective privacy-preserving feature extraction during interactions with people, but no statistically significant effects were found between the use of such privacy-preserving sensing, nor the robot’s transparency, and the breadth and depth of self-disclosure.

Persuasive social robots employ their social influence to modulate children’s behaviours in child-robot interaction. In this work, we introduce the Child-Robot Relational Norm Intervention (CRNI) model, leveraging the passive role of social robots and children’s reluctance to inconvenience others to influence children’s behaviours. Unlike traditional persuasive strategies that employ robots in active roles, CRNI utilizes an indirect approach by generating a disturbance for the robot in response to improper child behaviours, thereby motivating behaviour change through the avoidance of norm violations. The feasibility of CRNI is explored with a focus on improving children’s handwriting posture. To this end, as a preliminary work, we conducted two participatory design workshops with 12 children and 1 teacher to identify effective disturbances that can promote posture correction.

Maintaining a proper body posture during interactions with educational tablet applications is crucial for children’s physical well-being and task performance, especially considering digital tablet’s increasingly pervasive use in classrooms. In this work we propose WriteUpRight, an interaction system for children’s self-regulation of posture while writing on a tablet. The system relies on slowly deforming visual stimuli appearing on the tablet screen and compares two posture correction strategies: the Error Amplification method (see Figure 1) seeks to induce self-correction by amplifying the postural error, while the Error Correction method seeks to unobtrusively nudge the child towards the correct posture. Through a formative design and a user study with 42 children, we demonstrate the effectiveness of our solution and the advantages of the Error Amplification method with respect to the Error Correction method. The system shows potential for helping children maintain a proper head-screen distance and head roll angle during reading and writing tasks on tablets.

Social learning is a learning paradigm aiming to make the process of teaching new skills to an artificial intelligent agent (such as a robot) as close as possible to the process we employ when teaching skills to other humans. Equipping robot companions with the ability to learn from social interactions with their users would enable naive users to effortlessly teach skills to robots and support personalization and long-term use. Aiming to support research on the field, in this article we present a systematic literature review of human-like social learning for social robots, with a specific focus on works employing real robots and real human-robot social interactions. The review examines relevant papers from perspectives including the type of robot used, the task to be learned, and the method and assessment metrics employed, and allows for the identification of open research avenues.

Speech disfluency commonly occurs in conversational and spontaneous speech. However, standard Automatic Speech Recognition (ASR) models struggle to accurately recognize these disfluencies because they are typically trained on fluent transcripts. Current research mainly focuses on detecting disfluencies within transcripts, overlooking their exact location and duration in the speech. Additionally, previous work often requires model fine-tuning and addresses limited types of disfluencies. In this work, we present an inference-only approach to augment any ASR model with the ability to detect open-set disfluencies. We first demonstrate that ASR models have difficulty transcribing speech disfluencies. Next, this work proposes a modified Connectionist Temporal Classification(CTC)based forced alignment algorithm from [1] to predict wordlevel timestamps while effectively capturing disfluent speech. Additionally, we develop a model to classify alignment gaps between timestamps as either containing disfluent speech or silence. This model achieves an accuracy of 81.62% and an F1-score of 80.07%. We test the augmentation pipeline of alignment gap detection and classification on a disfluent dataset. Our results show that we captured 74.13% of the words that were initially missed by the transcription, demonstrating the potential of this pipeline for downstream tasks.

Communication between road users is crucial, as miscommunication and misunderstandings can lead to accidents, often bearing serious consequences in case Vulnerable Road Users (VRU) are involved. In this article we focus on communication solutions directed towards one particular group of VRUs, namely pedestrians. Various approaches for communication from vehicles to pedestrians have been proposed, including external Human-Machine Interfaces and Vehicle-to-Everything (V2X) communication. Surprisingly, these works are rarely looked at together. Therefore we jointly survey these separate fields as a starting point for the development of effective methods for interactions between vehicles and pedestrians. We argue that merging the perspectives of different fields can be beneficial as their approaches often complement each other.

Children’s retention of a proper body posture while interacting with educational tablet applications is important for both their physical health and task performance. In this work, we propose a new approach to unobtrusively induce postural changes in children by applying a slowly deforming visual stimulus appearing on the tablet screen. To preliminarily validate our approach we designed a reading-and-writing tablet application for children, during which 8 different slow visual stimuli would be provided, and monitored the children’s posture via a vision-based automated posture tracking system. Results from 10 children aged 6-11 suggest that the proposed approach is suitable for unobtrusively changing children’s postures and will stand as the basis for the future design of an adaptive unobtrusive posture regulation system.

Gender inequality is a widespread problem in our society. It can manifest itself in many ways and contexts, and starting as early as primary school. While an increasing number of initiatives aim at tackling gender biases and inequalities, few of them are aimed at raising awareness of gender (in)equalities among young children, i.e., at the age in which such inequalities appear in their lives. The potential shown by social robots in teaching non-curricular topics is a promising motivation for exploring their use in this context. Indeed, a social robot could offer children the possibility to discuss gender (in)equality with an intelligent entity that is neither male nor female, but rather a credible outsider with respect to mankind. In this article we present the design process of a social robot, named PixelBot, and associated robot-mediated storytelling activity aimed at raising awareness of gender (in)equality among children. We used a participatory design approach involving 20 children aged 1013 to acquire (i) their opinion on how a robot should look like and (ii) stories featuring robots and gender (in)equality. Finally, we conducted a study involving 8 children aged 9-10 to test the co-designed robot and robot-based storytelling activity. Results suggest that social robots are a promising avenue to promote gender equality and respect in children.

To make social robots effective in education, they need to be autonomous both in terms of assessing the student’s engagement state as well as intervening effectively in soft real-time when necessary. Hidden Markov Model (HMM) is an interpretable machine learning technique for modeling temporal data that is commonly used post-hoc to analyse latent learning processes. In this paper, we contribute by proposing an HMM-based intervention methodology for assessing and classifying the state of the student as either productive or unproductive in soft real-time. The system identifies and tracks states and patterns not conducive to learning, and a robot intervention is triggered whenever a too-high non-productive engagement is detected. In a pilot study with 22 children, we evaluate this methodology in terms of both 1) the effectiveness of the interventions on the students’ learning gains and on behaviors found conducive to learning, and 2) the students’ perception of the robotic interventions. Results suggest that the robot interventions have a positive effect on the post-test scores relative to the baseline robot, although there isn’t a significant difference in the learning gains. Moreover, interventions that try to induce reflective behaviors are most effective in inducing the required learning behavior, followed by communication-inducing interventions. Lastly, students’ perception of intervention usefulness does not reflect their actual effectiveness.

Handwriting practising, as any other repetitive task, often leads the practiser to an overconcentration state where their performance might be affected by postural and mental fatigue. Short breaks to perform unrelated activities, especially relaxation exercises, have shown to be a simple alternative to soften or postpone this phenomenon. Therefore, in this paper we are investigating the immediate effects of different types of short-duration relaxation exercises in the handwriting and posture of children aged from 8 to 10 in handwriting training. We divided 40 children in two groups performing the sessions, guided by a social robot, with small exercises of mindfulness or stretching in the middle of their training. Additionally, we analysed participants’ perceptions towards the robot leading these interactions. Results showed improvements in participants’ handwriting quality and posture maintenance regardless of the condition. Additionally, more positive feedback about the pause was reported from individuals in the mindfulness condition.

Context. With the introduction of Computer Science (CS) into curricula worldwide, teachers’ adoption of CS pedagogical content is essential to ensure the long-term success of reform initiatives. Continuing Professional Development (CPD) programs play a key role in this process. Unfortunately, adoption is seldom evaluated in CS-CPDs or CPDs in general. The result is a dearth of studies (i) modelling teachers’ adoption of CS pedagogical content or (ii) investigating factors influencing the uptake of this new discipline. Both aspects are crucial to design and characterize successful CPD programs. Objectives . We thus propose the Teachers’ Adoption of CS (TACS) model to investigate factors influencing the adoption of CS pedagogical content by teachers who are following a mandatory CS-CPD program. More specifically, the model proposes that contextual factors (e.g., age, gender, and general teaching experience), prior factors (e.g., experience, and CS perception), and acceptance factors (e.g., interest, and self-efficacy) may impact teachers’ adoption of CS pedagogical content. Methods. The study included 180 grades 5 and 6 teachers (students aged 9–11) that were following a mandatory CS-CPD program. The CS-CPD program involved participation in three-day-long sessions distributed over the 2019–2020 academic year. In between sessions, with the support of instructional coaches in the schools, teachers were encouraged, but not required, to adopt the CS pedagogical content. Therefore, during the CPD, and employing surveys based on the TACS model, we evaluated teachers’ adoption of the proposed content and investigated how the different factors influenced it. Results. At the PD-level, the results indicate that self-efficacy and interest queried during the CS-CPD are indicative of CS pedagogical content adoption. To shed more light on the relationship between these metrics, a more in-depth analysis was conducted with n = 92 teachers whose responses could be matched between sessions. While interest relates to how teachers adopt CS pedagogical content overall, both interest and self-efficacy are necessary to ensure the likelihood of a specific activity being adopted. Finally, individual teacher characteristics appear to impact adoption, with teachers with low experience with Information and Communication Technologies (ICT) requiring onboarding, while middle-aged teachers require convincing to adopt CS pedagogical content. Conclusion. Three takeaways emerge from the study. First, the analyses confirm the foundation of the TACS model. Second, the findings establish the key role that interest plays in said model. Finally, the results support the relationship between the contextual, prior and acceptance factors on the adoption of primary school CS pedagogical content.

Context Introducing Computer Science (CS) into formal education can be challenging, notably when considering the numerous stakeholders involved which include the students, teachers, schools, and policy makers. We believe these perspectives should be considered conjointly, which is possible within Research Practice Partnerships (RPPs) . RPPs look to bridge research-practice gaps and have seen an increase in the field of education and CS-education. Unfortunately, RPPs are considered to be under-researched, in addition to presenting their own challenges. Objectives To the purpose of assessing how RPPs may support the successful introduction of CS into formal education, we investigate three perspectives (students, teachers, and RPP stakeholders) and their interplay within the context of a multi-institution RPP conducting a pilot program to introduce CS to secondary school students. Methods A mixed methods analysis was employed to triangulate data in a concurrent triangulation design. The data included (i) three surveys distributed over the semester to 106 grade 9 students (ages 12-14), (ii) four teacher-journals, (iii) two interviews and four focus groups with the teachers and representatives of the partner institutions. Findings From the students’ perspective , while their self-efficacy increased, their motivation decreased throughout the semester due to a miss-match between their expectations and the course. The findings also indicate that gender biases and heterogeneity are already present in grade 9. From the teachers’ perspective , co-constructing the study plan, having access to regular support and collaborating within a community of practice when starting to teach CS all facilitated the teachers’ experience. Finally, from the RPP’s perspective the collaboration between stakeholders and having researchers evaluate the program were considered to be key elements in the pilot program. However, there appears to be a research-practice gap, in large part due to limited interactions between researchers and curriculum designers, and researchers and the teachers in the field. Conclusions From the students’ perspective it appears relevant to introduce CS (i) prior to secondary school to address motivation and bias-related issues early on, and (ii) to all students to avoid participation being motivation-, stereotype-, or belief-driven, and risk broadening the gap between students, (iii) all the while being attentive to course format and content to ensure that the course meets students’ expectations and fosters autonomous motivation. From the teachers’ perspective , while the support provided met the teachers’ needs, it is essential to find means of scaling such approaches when looking to deploy CS-curricular reforms to entire administrative regions. Finally, from the RPP’s perspective (i) teachers’ should be given a voice in the RPP to better align with the field, and (ii) researchers’ roles should be reconsidered to move beyond being only evaluators, and towards having a more co-constructive role in setting up the curricular reform. Recommendations are provided for researchers and practitioners involved in CS curricular reforms.

In this paper, we propose and compare personalized models for Productive Engagement (PE) recognition. PE is defined as the level of engagement that maximizes learning. Previously, in the context of robot-mediated collaborative learning, a framework of productive engagement was developed by utilizing multimodal data of 32 dyads and learning profiles, namely, Expressive Explorers (EE), Calm Tinkerers (CT), and Silent Wanderers (SW) were identified which categorize learners according to their learning gain. Within the same framework, a PE score was constructed in a non-supervised manner for real-time evaluation. Here, we use these profiles and the PE score within an AutoML deep learning framework to personalize PE models. We investigate two approaches for this purpose: (1) Single-task Deep Neural Architecture Search (ST-NAS), and (2) Multitask NAS (MT-NAS). In the former approach, personalized models for each learner profile are learned from multimodal features and compared to non-personalized models. In the MT-NAS approach, we investigate whether jointly classifying the learners’ profiles with the engagement score through multi-task learning would serve as an implicit personalization of PE. Moreover, we compare the predictive power of two types of features: incremental and non-incremental features. Non-incremental features correspond to features computed from the participant’s behaviours in fixed time windows. Incremental features are computed by accounting to the behaviour from the beginning of the learning activity till the time window where productive engagement is observed. Our experimental results show that (1) personalized models improve the recognition performance with respect to non-personalized models when training models for the gainer vs. non-gainer groups, (2) multitask NAS (implicit personalization) also outperforms non-personalized models, (3) the speech modality has high contribution towards prediction, and (4) non-incremental features outperform the incremental ones overall.

This article presents the novel modular version of the robotic platform Cellulo, a versatile handheld robot initially designed as an educational robot. The use of Cellulo in different contexts and applications over the years has highlighted the need for modularity. Modularity adds versatility by increasing the spectrum of functionalities of the robot, as well as more robustness. Modulo Cellulo consists of three modules: a main module, a battery module, and an interaction module. We describe the new Modulo Cellulo platform, the different modules design, the mechanical and electrical inter-connectivity between them, the new adaptive controller, and the application development framework. As a show case, we present the addition of the reconfigurable robot Mori as a module for Cellulo, in an activity envisioning the collaboration between reconfigurable swarm robots.

Wizard of Oz, a very commonly employed technique in human-robot interaction, faces the criticism of being deceptive as the humans interacting with the robot are told, if at all, only at the end of their interaction that there was in fact a human behind the robot. What if the robot reveals the wizard behind itself very early in the interaction? We built a deep wizard of Oz setup to allow for a robot to play together with a human against a computer AI in the context of Connect 4 game. This cooperative game interaction against a common opponent is then followed by a conversation between the human and the robot. We conducted an exploratory user study with 29 adults with three conditions where the robot reveals the wizard, lies about the wizard, and does not say anything, respectively. We also split the data based on how the participants perceive the robot in terms of autonomy. Using different metrics, we evaluate how the users interact with and perceive the robot in both the experimental and perceived conditions. We find that while there is indeed a significant difference in the participants willingness to follow robots suggestions between the experimental conditions as well as in the effort they put to prove themselves as humans (reverse Turing test), there isn’t any significant difference in their robot perception. Additionally, how humans perceive whether the robot is tele-operated or autonomous seems to be indifferent to the robot revealing its identity, i.e., the pre-conceived notions may be uninfluenced even if the robot explicitly states otherwise. Lastly, interestingly in the perception based conditions, absence of statistical significance may suggest that, in certain contexts, wizard of oz may not require hiding the wizard after all.

Robot-mediated learning activities are often designed as collaborative exercises where children work together to achieve the activity objectives. Although miscommunications and misunderstandings occur frequently, humans, unlike robots, are very good at overcoming them and converging to a shared solution. With the aim of equipping a robot with these abilities and exploring its effects, in this article we investigate how a humanoid robot can collaborate with a human learner to construct a shared solution to a problem via suggesting actions and (dis)agreeing with each other. Concretely, we designed a learning activity aiming to improve the computational thinking skills of children, in which the robot makes suggestions on what to do, that may be in line with what the human thinks or not. Furthermore, the robot may suggest wrong actions that could essentially prevent them from finding a correct solution. Via a pilot study conducted remotely with 9 school children, we investigate whether the interaction results in positive learning outcomes, how the collaboration evolves, and how these relate to each other. The results show positive learning outcomes for the participants in terms of finding better solutions, suggesting that the collaboration with the robot might have helped trigger the learning mechanisms.

Handwriting learning is a long and complex process that takes about ten years to be fully mastered. Nearly one-third of all children aged 4-12 experiences handwriting difficulties and, sadly, most of them are left to fight them on their own, due to the scarcity of tools for the detection and remediation of such difficulties. Building on state-of-the-art digital solutions for automated handwriting assessment and the training of specific handwriting-related skills, in this article we discuss requirements, rationale, and architecture of a system for handwriting training, which relies on a social robot as a mediator agent, offering personalized training and suggestions. The system is envisioned to operate autonomously and to support long-term interactions via personalization. Preliminary validation of the system in an experiment with 31 children showed its potential not only for autonomously guiding handwriting training sessions, but also for its inclusion in the teachers’ practice.

A wider incorporation of robots into classrooms is hampered by current technological limitations on full autonomy in social robots. Automated speech recognition, for example, a key enabler for vocal communication, is still unable to perform with sufficient accuracy. Past studies have shown that humans adjust their speech patterns to accommodate less skilled interlocutors. If such a response holds in human-robot interactions as well, we may be able to exploit it to lessen the burden on social robots and enable rich, autonomous vocal communication. In this paper we explore whether a robot’s speaking ability could have an impact on children’s speech patterns, learning, and engagement by designing an interaction where a child and a robot collaborate on a Tower of Hanoi puzzle. Sixteen children aged 7-14 completed this collaborative task partnered with a social robot that communicated with either high verbal (full sentences), low verbal (short phrases or single words), or nonverbal (sound-based utterances) vocalization. While we found no significant impact on children’s speech patterns or learning due to the robot’s method of vocalization, children in the non-verbal condition had a significantly lower perception of the robot’s intelligence along with higher rates of providing feedback and more instances of undoing its moves. This suggests that a link may exist between a robot’s perceived speaking ability and children’s confidence in that robot’s overall intelligence and capability in a collaborative task, as well as their empathy towards a peer they perceive as less skilled in the task.

Abstract Research has shown that Educational Robotics (ER) enhances student performance, interest, engagement and collaboration. However, until now, the adoption of robotics in formal education has remained relatively scarce. Among other causes, this is due to the difficulty of determining the alignment of educational robotic learning activities with the learning outcomes envisioned by the curriculum, as well as their integration with traditional, non-robotics learning activities that are well established in teachers’ practices. This work investigates the integration of ER into formal mathematics education, through a quasi-experimental study employing the Thymio robot and Scratch programming to teach geometry to two classes of 15-year-old students, for a total of 26 participants. Three research questions were addressed: (1) Should an ER-based theoretical lecture precede, succeed or replace a traditional theoretical lecture? (2) What is the students’ perception of and engagement in the ER-based lecture and exercises? (3) Do the findings differ according to students’ prior appreciation of mathematics? The results suggest that ER activities are as valid as traditional ones in helping students grasp the relevant theoretical concepts. Robotics activities seem particularly beneficial during exercise sessions: students freely chose to do exercises that included the robot, rated them as significantly more interesting and useful than their traditional counterparts, and expressed their interest in introducing ER in other mathematics lectures. Finally, results were generally consistent between the students that like and did not like mathematics, suggesting the use of robotics as a means to broaden the number of students engaged in the discipline.

Gamified rehabilitation tackles the problem of keeping patients engaged in, and motivated to do physical rehabilitation to improve its efficacy. However, with respect to standard rehabilitation, patients are freer to move about and may compensate their motion difficulties with parasite movements, which would greatly reduce the efficacy of the rehabilitation. To identify and characterize compensatory motions, we collected and analyzed video data of people playing the "tangible Pacman" game (an upper-limb rehabilitation game in which a patient moves a semi-passive robot, the "Pacman", on a map to collect 6 apples, while being chased by one or two autonomous robots, the "ghosts"). Participants include 10 healthy elderly adults and 10 chronic stroke patients, who played multiple runs of the game, with different sized maps and various game configurations. By analyzing the video recordings we successfully identified higher shoulder and torso lateral tilt compensation in stroke patients and developed a proof-of-concept compensatory motion detection system which relies on a wearable Inertial Measurement Unit and ROS to provide in-game, real-time visual feedback on compensation.

Collaborative learning appears in a joint intellectual efforts of individuals to understand an object of knowledge collectively. In their search for understanding the problems, meanings, and solutions, learners employ different multi-modal strategies. In this work, we explore the role of force feedback in learners interaction with tangible hand-held robots. We designed a collaborative learning environment to provide embodied intuitions on linear mathematical functions combined with graphical representations and ran a first study involving 24 participants. Our analysis shows a positive learning gain for our learning activity. Moreover, to explore the link between different types of force feedback and learners' collaboration, we designed a focus group study with 12 participants. Our results suggest that the haptic communication channel affects the collaboration dynamic differently according to the nature of the learning task. We finish by proposing design insights for future exploration of haptic in collaborative learning.

Humanoid robots, with a focus on personalised social behaviours, are increasingly being deployed in educational settings to support learning. However, crafting pedagogical HRI designs and robot interventions that have a real, positive impact on participants' learning, as well as effectively measuring such impact, is still an open challenge. As a first effort in tackling the issue, in this paper we propose a novel robot-mediated, collaborative problem solving activity for school children, called JUSThink, aiming at improving their computational thinking skills. JUSThink will serve as a baseline and reference for investigating how the robot's behaviour can influence the engagement of the children with the activity, as well as their collaboration and mutual understanding while working on it. To this end, this first iteration aims at investigating (i) participants' engagement with the activity (Intrinsic Motivation Inventory-IMI), their mutual understanding (IMIlike) and perception of the robot (Godspeed Questionnaire); (ii) participants' performance during the activity, using several performance and learning metrics. We carried out an extensive user-study in two international schools in Switzerland, in which around 100 children participated in pairs in one-hour long interactions with the activity. Surprisingly, we observe that while a teams' performance significantly affects how team members evaluate their competence, mutual understanding and task engagement, it does not affect their perception of the robot and its helpfulness, a fact which highlights the need for baseline studies and multi-dimensional evaluation metrics when assessing the impact of robots in educational activities.

A key element for the success of any game is its ability to produce a different experience at each round, thus keeping the player engagement high. This is particularly important for those games that also have a serious objective, such as gamified rehabilitation systems, aiming at encouraging patients in performing home rehabilitation exercises. In all cases, a game element which is typically static is the workspace, i.e. the "floor" upon which the game takes place. This is especially true for robot-assisted rehabilitation games, where the workspace must satisfy the requirements given by the robot's locomotion and localization systems, as well as the patient's exercise motion requirements.In this article we present a simple yet effective solution for designing dynamic and customizable tangible workspaces, which relies on hexagonal tiles and our previously proposed Cellulo localization system. These "hextiles" can be easily tangibly rearranged at each game round to yield a desired workspace shape and configuration, allowing tabletop mobile robots to move continuously within each new workspace. We ground our solution in the context of robot-assisted rehabilitation, where high adaptability is crucial for the efficacy of the solution, and propose a dynamic extension of our "tangible Pacman" rehabilitation game.Experiments show that the proposed solution allows for adaptation in range of motions, exercise types, physical and cognitive difficulty, besides reducing repetitiveness.

Personality is one of the most important factors in human interactions, which retains its importance in human- robot interactions with social robots. This work focusses on the varied linguistic strategies which characterize the personality traits of extraversion and introversion, analysing the main features that differentiate both personalities and eventually proposing a verbal communication model for the extraverted and introverted personality of a conversational social robot. The model classifies and converts phrases, with the result of building different communication styles. A pilot study, involving human subjects and the humanoid robot Pepper, programmed to mimic both extraverted and introverted personality types, has been conducted, with the twofold aim of assessing if differences between the two personalities of the robot can be perceived, and analyzing the effects of different personality traits on verbal interaction with human subjects. Preliminary results seem to confirm the law of attraction in human-human interaction for the extraverted personality.

The article focusses on the work performed in preparation for an experimental trial aimed at evaluating the impact of a culturally competent robot for care home assistance. Indeed, it has been estabilished that the user's cultural identity plays an important role during the interaction with a robotic system and cultural competence may be one of the key elements for increasing capabilities of socially assistive robots. Specifically, the paper describes part of the work carried out for the definition and implementation of two different robotic systems for the care of older adults: a culturally competent robot, that shows its awareness of the user's cultural identity, and a reference robot, non culturally competent, but with the same functionalities of the former. The design of both robots is here described in detail, together with the key elements that make a socially assistive robot culturally competent, which should be absent in the non-culturally competent counterpart. Examples of the experimental phase of the CARESSES project, with a fictional user are reported, giving a hint of the validness of the proposed approach.

The way in which humans behave, speak and interact is deeply influenced by their culture. For example, greeting is done differently in France, in Sweden or in Japan; and the average interpersonal distance changes from one cultural group to the other. In order to successfully coexist with humans, robots should also adapt their behavior to the culture, customs and manners of the persons they interact with. In this paper, we deal with an important ingredient of cultural adaptation: how to generate robot plans that respect given cultural preferences, and how to execute them in a way that is sensitive to those preferences. We present initial results in this direction in the context of the CARESSES project, a joint EU-Japan effort to build culturally competent assistive robots.

The functionalities and behaviours of socially assistive robots for the care of older people are usually defined by the robot's designers with limited room for runtime adaptation to meet the preferences, expectations and needs of the assisted person. However, adaptation plays a crucial role for the robot's acceptability and ultimately for its effectiveness. Culture, which deeply influences a person's preferences and habits, can be viewed as an invaluable “enabling technology” to achieve such level of adaptation. This paper discusses how guidelines describing culturally competent assistive behaviours can be encoded in a robot to effectively tune its actions, gestures and words. The proposed system is implemented on a Pepper robot and tested with an Indian persona, whose habits and preferences the robot discovers and adapts to at runtime.

Gestures are a natural communication modality for humans. The ability to interpret gestures is fundamental for robots aiming to naturally interact with humans. Wearable sensors are promising to monitor human activity, in particular the usage of triaxial accelerometers for gesture recognition have been explored. Despite this, the state of the art presents lack of systems for reliable online gesture recognition using accelerometer data. The article proposes SLOTH, an architecture for online gesture recognition, based on a wearable triaxial accelerometer, a Recurrent Neural Network (RNN) probabilistic classifier and a procedure for continuous gesture detection, relying on modelling gesture probabilities, that guarantees (i) good recognition results in terms of precision and recall, (ii) immediate system reactivity.

Modern manufacturing paradigms introduce the need for robots able to naturally cooperate with humans in an unstructured and dynamic environment. In this article we extend FlexHRC, an architecture for flexible and collaborative manufacturing robots, with an online perception-simulation-planning framework that allows the robot to assess the status of the workspace, keeping track at all times of the stage at which the cooperative manufacturing process is, to identify its next action, to simulate it to check its feasibility and, as a consequence, to dynamically allocate tasks to itself or the human operator. We have tested the FlexHRC with a dual-arm manipulator cooperating with a person to assemble a table with one tabletop and four legs.

Cultural adaptation, i.e., the matching of a robot's behaviours to the cultural norms and preferences of its user, is a well known key requirement for the success of any assistive application. However, culture-dependent robot behaviours are often implicitly set by designers, thus not allowing for an easy and automatic adaptation to different cultures. This paper presents a method for the design of culture-aware robots, that can automatically adapt their behaviour to conform to a given culture. We propose a mapping from cultural factors to related parameters of robot behaviours which relies on linguistic variables to encode heterogeneous cultural factors in a uniform formalism, and on fuzzy rules to encode qualitative relations among multiple variables. We illustrate the approach in two practical case studies.

In this paper we introduce a gesture-based robot control framework, we discuss the adopted design principles and we report results about its evaluation with humans. Gesture-based control using wearable devices may constitute a novel form of human-robot interaction, but its implications have not been discussed in the literature. We discuss the main challenging issues, possible design guidelines and an open source, freely available implementation using commercially available devices and robots. The overall performance of the architecture, as well as its validation with 27 untrained volunteers, is reported.

Cultural competence is a well known requirement for an effective healthcare, widely investigated in the nursing literature. We claim that personal assistive robots should likewise be culturally competent, aware of general cultural characteristics and of the different forms they take in different individuals, and sensitive to cultural differences while perceiving, reasoning, and acting. Drawing inspiration from existing guidelines for culturally competent healthcare and the state-of-the-art in culturally competent robotics, we identify the key robot capabilities which enable culturally competent behaviours and discuss methodologies for their development and evaluation.

Human Odometry (HO) is the process of providing a person with a continuous estimate of their location, on the basis of information acquired solely by sensors carried around by the person themselves. In an effort towards the development of effective and robust HO systems, we present the Human Odometry Outdoor Dataset (HOOD), a public collection of labelled accelerometer and gyroscope data recordings. We compare four sensor placements (foot, waist, wrist, chest) to identify the most suitable placement for different types of motions (ranging from walking to slithering), occurring in highly diverse real environments (such as flat grass fields, staircases and rough terrains).

Robots for the elderly are a particular category of home assistive robots, helping people in the execution of daily life tasks to extend their independent life. Such robots should be able to determine the level of independence of the user and track its evolution over time, to adapt the assistance to the person capabilities and needs. Human Activity Recognition systems employ various sensing strategies, relying on environmental or wearable sensors, to recognize the daily life activities which provide insights on the health status of a person. The main contribution of the article is the design of an heterogeneous information management framework, allowing for the description of a wide variety of human activities in terms of multi-modal environmental and wearable sensing data and providing accurate knowledge about the user activity to any assistive robot.

The automatic monitoring of specific Activities of Daily Living (ADL) can be a useful tool for Human-Robot Interaction in smart environments and Assistive Robotics applications. The qualitative definition that is given for most ADL and the lack of well-defined benchmarks, however, are obstacles toward the identification of the most effective monitoring approaches for different tasks. The contribution of the article is two-fold: (i) we propose a taxonomy of ADL allowing for their categorization with respect to the most suitable monitoring approach; (ii) we present a freely available dataset of acceleration data, coming from a wrist-worn wearable device, targeting the recognition of 14 different human activities.

The automatic assessment of the level of independence of a person, based on the recognition of a set of Activities of Daily Living, is among the most challenging research fields in Ambient Intelligence. The article proposes a framework for the recognition of motion primitives, relying on Gaussian Mixture Modeling and Gaussian Mixture Regression for the creation of activity models. A recognition procedure based on Dynamic Time Warping and Mahalanobis distance is found to: (i) ensure good classification results; (ii) exploit the properties of GMM and GMR modeling to allow for an easy run-time recognition; (iii) enhance the consistency of the recognition via the use of a classifier allowing unknown as an answer.

It is well known that there is a worldwide increase in both the number of elderly people and the number of elderly people with mild cognitive impairments [1], [2] and thus in need of assistance in the execution of activities of daily living. Socially Assistive Robotics is a novel research field, aiming at the design of robots relying on social means to interact with people and with a well-defined assistive purpose. The contribution of the article is three-fold: (i) a detailed analysis of the requirements of a socially assistive robot helping elderly people in the execution of everyday activities; (ii) the outline of the design principles for socially assistive robots; (iii) a first proposal for a wearable robot able to engage humans at the cognitive level.