ABSTRACT. Remote access technologies integrated with educational robotics environments provide students with revolutionary opportunities for robotics system engagement through cross-border, hands-on learning experiences. This paper presents pilot projects for remote access of a marine robot located in Germany from students in Germany, Croatia, Pakistan and Kazakhstan. Accessing robogics facilities located in a dif- ferent place may be the only opportunity for many students to operate robots - and in particular - marine robots - due to the challenges of hav- ing the right facilities in-situ. The feedback from the students showed that access barrier to technology was reduced. Students were enthusi- astic about the opportunity to learn more about and operate marine robots. Remote access has the potential to unlock new educational op- portunities, both in the formal degree-granting offering as well as in the lifelong learning, certificate education realm

ABSTRACT. This paper presents an innovative undergraduate course framework in cognitive robotics that combines hands-on engineering experience with theoretical foundations in sensorimotor learning. The ”Cognitive Robotics Project” course, offered in the third semester of a Humanoid Robotics program, enables students to design, construct, and program their own robots while understanding fundamental concepts of embodied cognition and sensorimotor exploration. The course consists of three parts: Hardware, Software, and Experiments. First, students solder custom-designed circuit boards and create 3D-printed robot designs. Then they implement sensorimotor behaviors using C programming. Finally, they conduct extensive experiments with their own hardware and software. The course’s unique approach not only teaches practical skills but also develops comprehensive understanding of the complete robotics development cycle. The course structure, technical framework, and educational outcomes are presented, demonstrating insights from the development and initial implementation of this course, including student-created robot designs.

ABSTRACT. The integration of robotics in education is transforming traditional teaching paradigms, presenting both opportunities and challenges. With emergence of the trend of practicing Robot Teachers as well boosting of AI tools, the investigators of this paper steered a small research study on Robot Teachers Vs. Human Teachers. The objectives of the said research were to study the comparative effectiveness of robot teachers and human teachers in fostering student engagement, learning outcomes, and emotional connection. Through a survey approach, data was collected from 533 students from various programs / disciplines as well from pre-service teachers. Quantitative analysis highlights the efficiency and consistency of robot teachers in delivering standardized content, while qualitative findings underscore the irreplaceable role of human teachers in nurturing empathy, creativity, and personalized guidance. The study reveals that while robot teachers excel in repetitive and data-driven tasks, human teachers remain pivotal for fostering critical thinking and emotional intelligence. A hybrid model, leveraging the strengths of both, emerges as a promising approach to future education. This paper concludes by discussing implications for policy, curriculum design, and teacher training, advocating for a balanced integration of robotics to enhance educational outcomes without compromising the human touch.

ABSTRACT. Computational thinking (CT) is regarded as a crucial skill that every 21st-century citizen should possess. Therefore, the development of this skill should start at preschool age. According to the existing literature, educational robotics may be a vehicle for CT enhancement. However, despite the increased interest of researchers in implementing programs that contribute to the development of computational thinking in educational environments, empirical research in the field of preschool education is limited. The aim of the pilot study was to examine the potential role of educational robotics in fostering the development of CT among preschool-aged children and the gender and age effect. This study used structured problem-based robotic activities developmentally appropriate for 4-6 aged preschoolers to cultivate CT while programming the robot Beebot. Activities implementing young children in simple and complex algorithms, decomposition, pattern recognition, and debugging were used. Results indicated that a robotic course could promote overall CT development in kindergarten settings. Gender and age factors did not affect kindergartener's CT scores. Researchers in the field of CT may benefit from this work regarding the best tools for CT development.

ABSTRACT. How do different designs of an Educational Escape Room (EER) facilitated by social robot influence learning outcomes and game experience? This paper presents the results of a randomized control study on the use of the social robot (SR) NAO as a facilitator in a cooperative and competitive EERs designed to enhance students’ knowledge of AI and machine learning (ML). The study involved 52 undergraduate students who were assessed on learning outcomes and game experience through pre- and post-surveys. The results show that both EERs led to significant improvements in AI/ML knowledge, with cooperative mode enhancing competence and immersion, while competitive mode increasing flow, challenge, and positive affect. These findings demonstrate the potential of EERs facilitated by SR to foster learning outcomes and positive game experience in AI education, providing insights into differential effects of cooperative and competitive modes in EERs with SR.

ABSTRACT. Nowadays, new technologies and educational robotics are regarded as contributors to effective learning from an early age and as useful tools for educators. Yet, the educational robotics trends, particularly in domains, such as mathematics in early childhood education (up to 6 years old), have not been thoroughly investigated and systematically presented. Therefore, this current work employs bibliometric and content analysis to map and explore the field of educational robotics related to preschoolers’ mathematical skills. In detail, 36 papers retrieved from the Scopus database have been studied thoroughly and analyzed with VOSviewer and Atlas.ti software. Our findings show that the USA has the highest publication performance in the related field. Bers and Sullivan arise as the most influential authors. Moreover, we reveal that researchers often choose the quantitative method to analyze data stemming from experimental studies mainly. Furthermore, tangible robots (BeeBot and KIBO) are frequently used in STEM (Science, Technology, Engineering, Mathematics) and computational thinking activities correlated with Maths without though focusing on other skills such as the 4Cs (communication, collaboration, critical thinking, creativity). In addition, robots are used solely as programming tools. Participants were asked to navigate them on mats to achieve specific goals during the intervention Besides, warm-up and free experimentation activities are limited. Accordingly, we suggest designing robotics activities tailored to kindergarten age and needs to enhance various mathematical skills. Finally, step-by-step activities within learning models suitable for preschool such as the 5E model are recommended.

ABSTRACT. Current AI education concepts often emphasize supervised machine learning algorithms, while reinforcement learning (RL) education remains limited, typically focusing on introductory concepts. To address the need of a more in-depth RL educational experience, we propose two hands-on robotic activities to introduce foundational RL concepts and foster exploration of RL system design principles. These activities include a LEGO car task and a beam-and-ball balance robot simulation. We conducted a pilot study featuring this two-part curriculum with 13- to 16- year old students and analyzed their learning outcomes quantitatively. Our findings indicate that middle and high school students developed an understanding of basic RL concepts as measured by post-activity reflections. We also discuss next steps to enhance the curriculum, including providing a more interactive experience with RL system design.

ABSTRACT. This work examines the role of educational robotics towards teaching high-school children basic automatic control concepts including feedback, the structure and the differences between open- and closed-loop control systems, the basic elements of a closed-loop system, the effect of external disturb-ances, the tracking error, as well as practical issues concerning control hard-ware implementation. The ideas are exemplified with hands-on exercises and experiences from a STEM summer school based on educational robotics.

ABSTRACT. Short paper about the outdoor robot competition, its rules and experiences. Also there are three team robots and their HW and SW equipment described.

ABSTRACT. The international FIRST® LEGO® League (FLL) Challenge robotics competition is designed to promote students’ STEM interests and experiential development of 21st Century skills. This paper reports on the findings of a qualitative interpretivist multiple case study PhD project which examined the lived experience of FLL Challenge robotics coaches and student teams in Perth, Western Australia. It compares students’ self-reported perceptions of their 21st Century collaboration and communication skill development with their actual learning and team processes during the 2021 competition season, as observed by their coach and a non-participant researcher. Students reported positive impacts of competition participation on their team’s skill development; however, coach and researcher observations suggest that social and teamwork challenges meant that these improvements were not necessarily demonstrated to the same extent by all team members. Students’ skill development during the season was influenced by individual team members’ expertise, skills, and attitudes; and their coaches’ facilitation of reflective team building tasks.

ABSTRACT. This paper presents a software and hardware framework for using slot cars as a competitive learning and developmental platform in software engineering. A novel mathematical model of the slot car is presented, facilitating experimentation and development of control algorithms and tracking systems in a simulated environment. Additionally, we present a minimally invasive modification to a Carrera slot car track which seamlessly transitions the algorithms from simulation to a real world setup. The slot car's in-track states are estimated using an extended Kalman filter and a web camera, with flexibility for integration with other sensors. The Kalman filter gives real-time tracking of the slot car states, allowing for control feedback during the race. A comparison between physical trials and simulator runs are shown to closely match. The low dependence on hardware modification and relative low cost makes this an excellent choice for a hands-on capstone project or competitive programming challenge.

ABSTRACT. The Sphero Bolt robot is an educational tool designed to make learning about coding and robotics engaging and hands-on. How- ever, existing software environments present challenges for classroom use, particularly due to limited support for open platforms. We developed a free, browser-based IDE for the Sphero Bolt hardware. It is built on top of Node.js and WebAssembly, using Pyodide. The IDE enables direct in- teraction with the robot not requiring additional software installations. This paper outlines the technical and educational contributions of our IDE, including an example workshop on behavioral control concepts with practical programming exercises. Inspired by Valentino Braitenberg’s ve- hicles, we demonstrate how students can use the robot’s light sensors to simulate reactive behaviors such as light avoidance and attraction. By removing technical barriers and providing a seamless portable coding environment, this IDE lets students and educators focus on the creative and exploratory aspects of programming and robotics.

ABSTRACT. This paper discusses how AI-powered educational robotics tools can enhance students’ understanding of AI, promote AI literacy, and support their effective and ethical use of AI with their robotics projects. It introduces AI literacy and various frameworks that support student’s AI learning. It bridges AI literacy, AI content knowledge (Five Big Ideas in AI), and pedagogical approaches and proposes an AI Literacy Framework for AI-embodied Educational Robotics. The framework aims to support educators in providing effective AI-powered educational robotics learning opportunities to enhance their AI learning through hands-on robotics-making activities.

ABSTRACT. This study investigates the impact of activities using an AI-driven robot on the metacognitive behaviors and mathematical problem solving performance of primary school students. In collaboration with four teachers, we developed a pedagogical sequence combining educational robotics learning activities that embed an AI algorithm (AI-ERLA) with transfer tasks in mathematical problem solving. Qualitative and quantitative results, based on teacher observations and pre-post math tests (N=85), demonstrate positive effects of AI-ERLA, with three significant findings: (1) change in students' attitudes toward error, promoting active engagement; (2) improvement in problem-solving strategies, with more reflective strategies and analytical thinking; and (3) statistically significant increase in overall performance, particularly among low-achieving students and those with special needs. The results highlight the potential of AI-ERLA to improve students' metacognitive skills and mathematical problem solving. While promising, further experimental studies with larger samples are needed to validate these initial findings.

ABSTRACT. Robot assistants offer a variety of benefits in higher education classrooms, including addressing the shortage of teachers and increase of student engagement. However, due to their high price, the robots should be able to per-form various tasks to fully utilize their potential. We tested the TEMI V3 robot assistant, driven by the custom-developed Teaching Assistant application, in three key classroom scenarios: attendance tracking, hybrid learning, and micro-learning. Nine students from Tallinn University of Technology evaluated the robot’s features and gave their feedback via semi-structured interviews. Our findings suggest that while teachers value robot assistants’ help with tedious routine tasks such as attendance check, the robots are also beneficial for hybrid learning by providing students with physical and social presence, and can be somewhat helpful for catching up with lesson contents. However, issues with privacy, authenticity of attendance records and technical questions were also brought up.

ABSTRACT. The teaching of engineering and computer science often relies on theoretical instruction with limited practical applications. This lack of hands-on experience can reduce student engagement and comprehension, contributing to high dropout rates in technical courses. To address this issue, we developed Baratinha, a compact and modular educational robot designed to integrate theory and practice in university classrooms.

Baratinha is a 10 cm diameter two-wheeled mobile robot, powered by an ESP32-S3 N16R8 microcontroller, and equipped with infrared reflectance sensors, encoders, an inertial measurement unit (IMU), and Time-of-Flight (TOF) distance sensors. It supports WiFi and Bluetooth communication, enabling multi-robot interconnectivity and IoT applications. The robot is programmable in C++ (Arduino Framework, VS Code), MicroPython and ESP-IDF, offering flexibility for different educational contexts.

This paper presents the hardware and software architecture of Baratinha and explores its applications in control systems, robotics, artificial intelligence, and embedded systems. The robot enables practical experiments, such as PID control, autonomous navigation, swarm robotics, and real-time sensor data acquisition, making it a versatile tool for hands-on learning. Preliminary feedback from students—many of whom have excelled in international robotics competitions—indicates that Baratinha significantly enhances engagement and comprehension. Future developments aims to enchance computational capabilities, broaden sensor integration for AI applications, and refine software documentation for enhanced usability. Through this interactive platform, Baratinha seeks to advance engineering education and promote hands-on student engagement.

ABSTRACT. This article presents aiMetalSorter, an educational project aimed at raising awareness about waste generation and exploring strategies to achieve zero waste. Using the micro:bit board, the students developed a system that combines sensors, programming, and artificial vision to classify recyclable materials. A key focus of the project was understanding the concept of sensors. The students built a metallic detection sensor using two separated wires that detect a change in electrical current flow when a metallic object connects them. They also learned programming by designing and coding the control system for a servomotor that operates a robotic arm to sort the materials. Furthermore, the project introduced students to image recognition through artificial vision by programming the micro:bit to identify the shape of metallic objects (circular or square). This hands-on experience provided an integrated approach to learning about sensors, coding, and vision-based systems while fostering critical thinking about sustainability and the circular economy.

ABSTRACT. This research revealed the views and experiences of STEAM field teachers on the flipped learning approach and examined the effects of this method on their robotics and coding domain knowledge. Mixed research method including both qualitative and quantitative methods were used together in the study. Eleven STEAM teachers working in public schools in Bursa, Turkey participated in the study. Qualitative data were collected through semi-structured interviews with teachers and evaluated by descriptive analysis method. Quantitative data were collected through a test prepared to measure robotics and coding achievement and the data was analyzed by Wilcoxon signed ranks test. In line with teachers' opinions, research findings showed that teachers generally wel-come the flipped learning method, the environment supported individual learn-ing and increased classroom interaction. In addition, teacher opinions empha-sized that flipped learning encourages active learning by increasing the self-confidence of students and teachers. However, it has been determined that fac-tors such as teacher attitudes, lack of equipment, high workload and students who do not have learning responsibility may make it difficult to implement the flipped learning method. Quantitative analyses revealed that the flipped learn-ing approach significantly increased teachers' academic achievement in robot-ics and coding. As a result, the integration of the flipped learning approach with STEAM education is seen as an effective method, especially in applied courses. However, in order for the method to be applied efficiently, it is neces-sary to strengthen the infrastructure and increase the support of teachers.

ABSTRACT. Understanding user intentions through text is a core issue for conversational robots in the education field. However, the students’ posts, responses, and chats for chatting robot or chatbot systems are user-generated content (UGC) that generally cannot be moderated. A large number of Chinese homophones cause confusion for user intent recognition. Therefore, the chatbot may not be able to obtain accurate user intent due to misunderstandings by the speech recognition system, user typing errors, user phonetic pinyin errors, or spelling errors. However, in the literature, most generation- or retrieval-based Chinese chatbot systems ignore these situations. This study proposes the Chinese Similar Synonyms Model (CSSM), which builds Chinese similar synonym knowledge bases with deep learning to solve the UGC problem of Chinese conversational robots. Similar synonyms are words that have the same function as synonyms but are expressed in a different way. Similar synonyms may appear due to misunderstandings in the Chinese speech recognition system, homophones, incorrect associations in the Chinese auto-correction or auto-complete mechanisms, or user input errors. Similar synonyms in Chinese may prevent Chinese chatbots from recognizing the exact word and user intent. We established four knowledge bases to handle similar synonyms in Chinese, including morphologically similar synonyms, adjacent similar synonyms, homophonic similar synonyms and confusing phonetically similar synonyms. Two word embedding methods are used, namely Keras, Word2vec Skip-gram and two extended versions of recurrent neural networks, namely long-short term memory (LSTM), gated recurrent unit (GRU). We test our CSSM for four Chinese question-answering domains and experiments show consistent positive results across these domains.

ABSTRACT. Prolonged school absences due to chronic illness significantly compromise children and adolescents' sense of belonging, a fundamental aspect of their well-being. To address the resulting social isolation and potential academic, social, and emotional challenges, telepresence robots, specifically the Avatar AV1, have emerged as a promising intervention. This research explored the impact of Avatar AV1 on school participation, social inclusion, and overall well-being among students with chronic health conditions. A mixed-methods approach, combining qualitative interviews (with students, parents, teachers, and classmates) and quantitative data (measuring Avatar AV1 usage patterns and students’ reported sense of school belonging and participation), was employed. Qualitative data were analyzed using thematic analysis, while quantitative data were analyzed descriptively. The integration of these datasets facilitated a comprehensive understanding of the Avatar AV1 experience. Key success factors and crucial moderating influences are identified and discussed in the presentation.