ABSTRACT. This paper investigates whether systematic curiosity-eliciting questioning strategies induce epistemic curiosity in students and enhance learning outcomes in robot-assisted pitch training. Embedded in a project-based media course where students developed AI solution concepts for industry partners, the training aimed to support students in preparing their final pitch presentations. The Furhat robot, integrated with a Large Language Model, served as an interactive pitch coach. Students were randomly assigned to one of two conditions: curiosity-eliciting questioning (n=14) or standard constructive feedback (n=17). The curiosity condition implemented six theory-based strategies (counterfactual questions, paradoxical challenges, comparative questions, perspective-taking, future projections, and provocations), while the control condition received tra-ditional supportive feedback. Results from our randomized controlled trial showed statistically significant advantages for the curiosity condition on pitch skill improvement (d=0.77, p=.043) and conceptual insights (d=0.97, p=.012). Results also showed that the curiosity condition significantly increased epis-temic curiosity (d=0.75, p=.048). Qualitative data triangulated these findings: 85.7% of curiosity participants versus 17.6% of controls identified critical questioning as curiosity-inducing. Correlation analysis confirmed that experi-enced curiosity significantly predicted learning outcomes (r=.65, p=.012). Furthermore, students in the curiosity condition demonstrated consistent im-provements with substantially lower variance in conceptual insights (SD=0.54) compared to controls (SD=1.15). These findings demonstrate that social robots can effectively elicit epistemic curiosity as a learning mechanism and that sys-tematic curiosity-eliciting strategies may lead to improved outcomes in educational contexts.

ABSTRACT. Dealing with complex socio-technical systems requires future skills that go beyond technical and digital competencies. The \textit{Future Skills 2030} approach focuses on community-oriented competencies and ambiguity competence. While these skills are increasingly required in industry and the labor market, they have not yet been systematically incorporated into higher education.

Robo4Ever addresses this discrepancy with a research-oriented teaching and learning format. Students from different disciplines work in interdisciplinary teams on application scenarios of their own choosing for various assistance robots. In doing so, they develop technical, social, and ethical frameworks, negotiate disciplinary perspectives, and reflect on the collaborative work process. Robotics serves as a socio-technical learning medium that combines technical feasibility with social impact.

Graphically combinable basic robot functions reduce technical barriers to entry and shift the focus to collaborative problem solving and reflection.

An exploratory focus group interview shows that students measure learning content less by its depth and more by how responsibly knowledge is imparted, applied collectively, discussed, and reflected upon. Reflection rather than rote learning forms the core of Robo4Ever.

ABSTRACT. Whatever it concerns young learners or older ones, Educational Robotics and Computational and Algorithmic Thinking (CAT) share common target competences. The case of training teacher students of a master in educational sciences is illustrated. This course (Introduction to CAT) exploited resources produced through a participatory design process during the EU Erasmus+ PIAF project dedicated to the development of CAT. A basic principle is observed in the design of learning scenarios: respecting pedagogical alignment between target competences, learning activities and evaluation. Field testing of educational scenarios with the target publics and their adaptation offer sharable resources (potentially adaptable) by other « trainers » (teachers, animators). A particular attention is dedicated to a roadbook use to enhance metacognition and follow the collaborative design of an educational scenario by our students and to observe the acceptability of the PIAF resources. Recom-mendations are provided about trainers’ training, competences, learning methods and evaluation, introduction in curricula.

ABSTRACT. This study examines the attitudes of pupils ages 10-13 with regard to their interest in scientific disciplines involving robots and technologies, the career opportunities these fields offer, and their personal career aspirations. The main aim is to investigate whether the participation in an annual school-based STEAM laboratory, featuring Educational Robotics (ER) activities, has an impact on the attitudes of pre-adolescent pupils toward science. The research was conducted in a school with an emphasis on inclusion for children with special needs. Research indicates that ER can promote inclusion in schools by addressing differences in learning, disability, and gender (Belpaeme et al., 2018; Kucuk & Sisman, 2020; Lehmann, 2020). As part of a qualitative intervention study, 41 students completed anonymously pre- and post-test questionnaires to measure potential changes. The data were analyzed using purposive non-probabilistic sampling: the results show that general attitudes remained largely unchanged between pre- and post-laboratory surveys. However, we found a significant increase in motivation to participate in scientific and technological activities in female students, highlighting the potential of STEAM laboratories to foster engagement and gender equality. These findings emphasize the usefulness of STEAM laboratory experiences in schools and encourage further research on their impact on both students’ attitudes and learning outcomes in ER-based activities.

ABSTRACT. This article explores the impact of hands-on educational robotics on preschool age children and the skills that can be developed through interaction with robots and themed activities. It covers the structure of the activities and the feedback received from initial testing. The experimental study focuses on five activities, all of which were designed to involve robots. The results are presented in terms of five skills, focusing on concentration, involvement in each activity, initiative, manner of expression and cooperation during the activities. The effectiveness of educational robotics at this developmental stage is investigated focusing on its contribution to skill development.

ABSTRACT. FIRST LEGO League (FLL) is a long-running and widely adopted educational robotics competition. While prior research has examined learning outcomes and participant experiences, little attention has been paid to how the technical skills emphasized within the competition itself evolve over time. This study adopts a program-centered perspective to analyze shifts in navigation-related technical skill emphasis within FLL across three generations of LEGO educational robotics platforms—NXT, EV3, and SPIKE Prime—spanning twenty competition seasons from 2006/2007 to 2025/2026. Focusing on the Robot Game and Robot Design areas, the analysis examines judging rubrics, robot game field designs, and platform capabilities to identify how navigation-related skills are required and evaluated across hardware generations. Survey data from recent seasons provide contextual insight into contemporary practices. The results show a redistribution of technical demands rather than a reduction: low-level navigation control becomes increasingly abstracted, while strategic planning, systematic testing, and iterative refinement gain prominence. These findings highlight competition design as a mediating mechanism shaping how technical skills are made visible and valued in educational robotics programs.

ABSTRACT. Educational robotics is widely recognized as a powerful approach for fostering active learning, problem-solving, and engagement in primary education. However, while robotics workshops generate rich learning situations, their potential as contexts for structured observation of learning-related behaviors remains largely underexplored. This paper presents the design and iterative content validation of an observation-based tool embedded in primary school robotics workshops, aimed at supporting early pedagogical alert regarding learning difficulties.

Grounded in established frameworks from developmental psychology and cognitive neuroscience, the tool targets observable behaviors across four developmental domains : language, motor, socio-emotional, and cognitive, while considering executive and cognitive functions as underlying explanatory mechanisms. The tool was developed through a staged, theory-driven process involving iterative design, field experimentation, and expert consultation, culminating in a Delphi-based content validation phase. Practitioner feedback played a central role in ensuring feasibility and observability within real workshop constraints, leading to a final 20-item version usable by non-psychologist robotics facilitators.

Rather than aiming at diagnosis or psychometric assessment, the proposed approach supports early pedagogical alert by enabling structured observation of learning-related behaviors in authentic, activity-based contexts. The paper argues that educational robotics workshops constitute privileged environments for such observation and discusses the methodological scope, limitations, and future validation pathways of the tool.

ABSTRACT. This paper explores the implementation of educational robotics as a tool for physically visualising Lissajous figures, a family of complex curves generated by the intersection of two perpendicular harmonic oscillations. Lissajous figures have interesting scientific history implications and have many applications in engineering and physics. These patterns are traditionally studied using computer simulations or oscilloscopes, which methods have a lower pedagogical impact because they obscure the physical mechanisms by which the curves are generated. This study proposes a practical pedagogical approach that uses a programmable robot equipped with a drawing tool to draw Lissajous figures on a sheet of paper. The paper tray and the pen move perpendicular to each other, both performing harmonic motion, achieved using a Scotch yoke mechanism. The parameters that determine the behaviour of the curves can be selected from a wide range, allowing students room to experiment. The relationship between settings and visual output is easy to understand, but studying how the robot works reveals deep connections between physical, mechanical, and mathematical phenomena.

ABSTRACT. Shoebox tasks are widely used in the education of individuals with Autism Spectrum Disorder (ASD) to support learning through clear visual structure. Previous work demonstrated adapting shoebox tasks for robot-assisted education, with a robot acting as a teacher using automated task analysis and action planning via a vision-based solver. In this paper, we extend the solver to a new use case: teaching programming and robotics. We present a system combining shoebox tasks, an enhanced task solver, and a visual robot interface to support programming learning at multiple levels of abstraction. The solver and interface structure student programming activities through card-based visual programming, allowing students to construct robot behavior without code by mapping card sequences to executable programs. Here, a sequence of visual cards represents an algorithm, reflecting the use of visual schedules in the Treatment and Education of Autistic and Related Communication Handicapped Children (TEACCH) methodology for visualizing daily activities. The system requires no prior programming skills and is applicable to different robot platforms, including humanoid robots and low-cost educational floor robots. Its design aligns with structured teaching practices in ASD education, supporting inclusive and accessible robotics learning. Classroom examples demonstrate applications using the Nao humanoid and Ozobot educational robots.

ABSTRACT. This quasi-experimental study assessed how insect-robot competitions, used as an innovative didactic strategy, impacted elementary school students in a rural context. Researchers worked with 31 sixth-grade students (ages 10-12) attending Simón Rodríguez School in Riobamba, Ecuador. The interven-tion guided students in constructing and competing with insect robots made from recyclable materials and essential electronic components. Pre-post test analysis revealed students made significant improvements across four di-mensions: technical comprehension (Δ=2.38), practical skills (Δ=2.21), atti-tudes and motivation (Δ=2.34), and specific capabilities (Δ=2.33). Statistical analysis identified significant differences (p<0.001) between pre- and post-intervention measurements. These findings indicate that educational robotics competitions effectively develop STEAM competencies in rural contexts with limited resources, supporting meaningful learning and technological in-clusion, evidence supporting the positive impact of insect-robot competitions as a didactic strategy in rural education. The findings indicate significant im-provements across multiple learning dimensions, with effect sizes that sur-pass those reported in urban contexts. These findings extend the existing ev-idence to rural contexts, demonstrating that effective technological strategies are feasible even with limited resources.

ABSTRACT. The industrial robotics sector faces a significant workforce skills gap, with traditional semester-long courses struggling to meet industry's urgent need for professionals skilled in modern tools like ROS 2. This study presents a qualitative evaluation of a two-week intensive industrial robotics summer school designed to rapidly develop practical competencies. Using thematic analysis of daily learning diaries (n=10) complemented by pre-post confidence surveys (n=9), we examined how students experienced and progressed through compressed technical education. Five themes emerged: technical skill breakthroughs, problem-solving evolution, collaborative learning dynamics, theory-practice integration, and self-efficacy development. Findings reveal that intensive formats can effectively develop industry-relevant robotics skills through hands-on engagement, progressive complexity, and authentic industry exposure. The study provides practical recommendations for designing intensive robotics courses that address workforce development needs.

ABSTRACT. Bridging the gap between industrial robotics standards and academic curricula presents significant pedagogical challenges. This study evaluates the transferability of an advanced ROS 2 navigation workshop across general, professional, and academic audiences. Using a quasi‑experimental design, we assessed learning outcomes through pre‑ and post‑workshop self‑evaluations, cognitive load using NASA‑TLX, and satisfaction using both SEEQ and an ad‑hoc technical questionnaire.

Results showed that learning gains and cognitive load remained stable across technical cohorts, demonstrating that complex industrial material can be adopted in higher‑education settings without increasing cognitive burden. However, satisfaction varied significantly; voluntary attendees (general and professional) reported higher levels of enthusiasm and more positive attitudes toward the instructor compared to the mandatory academic cohort, despite comparable learning improvements.

Follow‑up analysis revealed that this divergence stemmed not from the technical infrastructure but from differences in expectations regarding onboarding and procedural scaffolding. These findings suggest that while the technical content is transferable, success in academic settings requires adapting industrial training to provide more structured support.

ABSTRACT. In 2024 EduRob was introduced as a versatile and highly reconfigurable educational robot specifically designed for teaching the kinematics of wheeled mobile robots. The small modular robot has gone through a major hardware revision and is successfully being incorporated into multiple bachelor and master modules at the depart- ment of computer science, electrical engineering, mechanical engineering and information technology at the University of Applied Sciences and Arts in Dortmund. Offering the possibility to switch between different kinematics including standard wheels, Omni wheels and Mecanum wheels in various configu- rations, students can test their theoretical understanding of transforma- tions and control in mobile robotics. Through the integration of micro-ROS into EduRob’s firmware high-level ROS-based robot software stacks can be implemented especially useful for the curriculum of master’s programs.

ABSTRACT. The use of simulation has become a vital component of robotics education in academia. Bringing physical robots into the classroom is often impractical due to physical setup constraints, hardware size, and logistical limitations. Furthermore, for students attending lectures online, these challenges are even more pronounced. As a result, lecturers increasingly prefer simulation environments for teaching robotics concepts. However, simulation-based learning environments also introduce several challenges. These include the need to switch to Linux as the operating system, additional installation and configuration steps when using VirtualBox or Docker-based solutions, and the lack of powerful GPUs on student devices required for computationally intensive 3D simulations. Moreover, individual installation errors frequently require lecturer support, and significant classroom time is often spent guiding students through software and dependency installation.

To overcome these limitations, we developed RoboEdu, a remotely accessible web-based environment that facilitates a ROS 2 learning environment with all required software, packages, and course repositories pre-installed, providing students with a ready-to-use workstation. The RoboEdu learning environment is de- ployed on a virtualized Linux server with full GPU acceleration. It creates iso- lated ROS 2 environments using Docker containers, and access is provided via noVNC secured with strong authentication. The ROS2 environment includes simulation tools such as Gazebo and RViz, along with all necessary supporting software. In earlier versions of RoboEdu, users experienced graphical flickering due to improper GPU acceleration within containers. This issue has been resolved in the current implementation, which now fully utilizes GPU acceleration.

ABSTRACT. Telepresence robots (TPRs) are used in higher education when a teacher cannot be physically present. Previous studies suggest that, compared with conventional videoconferencing-based hybrid learning, TPR-mediated teaching can enhance students’ sense of social presence. However, less is known about how students experience TPR-mediated teaching across different cohorts and instructional contexts. This study examines students’ perceptions of TPR-mediated teaching, focusing on communicative interaction, content understanding, and perceptions of spatial placement and safety. The PEPCII pedagogical design model was used as an analytical framework to structure and compare experiential components of embodied hybrid learning. Data were collected through a mixed-methods design, including a questionnaire with Likert-scale items and open-ended responses, and analysed using descriptive and comparative statistical analysis as well as qualitative content analysis to examine cohort- and context-related differences. The results indicate that students in both cohorts evaluated teacher communication and content understanding positively. The spatial placement of the robots was perceived as a stable experiential component, whereas communicative clarity was more sensitive to contextual variation across cohorts, while content understanding remained consistently high. Student engagement was mixed but generally sustained, with interaction patterns shifting toward more structured participation. Overall, the findings highlight that the effectiveness of TPR-mediated teaching is shaped by pedagogical alignment, interaction norms, and classroom context, sup-porting the conceptualisation of TPR-mediated teaching as a socio-pedagogical arrangement rather than a fixed technological solution.

ABSTRACT. In this paper, we present our experiences and lessons learned from conducting a series of global virtual classroom (GVC) courses on robotics. The main focus of these courses was to provide students from different universities with an opportunity to work in international teams on challenging robotics projects. We describe the course structure, the projects that we used to motivate the students: a Bamboleo playing robot, a TurtleBot3 obstacle run, and a rock climbing humanoid robot. Our evaluation indicates that students benefited significantly from the collaborative learning environment, gaining practical skills in teamwork and robotics. We discuss important factors for the success of the GVC courses: team formation, ice breaking. class scheduling, and assessment.

ABSTRACT. This two-stage iterative study conducted across two academic years evaluates the effectiveness of TPRs for remote teaching in a university course by comparing student engagement across multiple instructional modalities, including telepresence-based teaching and conventional videoconferencing. Student feed-back highlights that increased mobility and enhanced sense of presence afforded by TPRs improved engagement, and offered greater flexibility for teachers, ad-dressing key limitations of teleconferencing platforms. Overall, the findings indi-cate that TPRs can substantially improve social presence, remote teacher mobility, and individualized support.