Introduction

Learning difficulties remain a persistent concern in various areas of knowledge because they shape the academic trajectories of students and are often insufficiently addressed in everyday classroom practice. Despite the rapid growth of educational robotics as a promising tool for active and inclusive learning, students with learning difficulties remain underrepresented in educational robotics initiatives, and prior studies rarely report concrete strategies for their meaningful participation in school settings. This gap motivates our study, as we investigate how educational robotics can be integrated into pedagogical practice to support the learning process of students with learning difficulties, presenting the challenges, limitations, and opportunities observed in public schools.

Multiple interrelated factors can contribute to the emergence of learning difficulties and may include emotional and social conditions (e.g., anxiety and unstable family environments), specific neurocognitive conditions (e.g., ADHD, dyslexia, and dyscalculia), unaddressed learner variability, health constraints (visual or hearing impairments and chronic conditions), adverse environmental conditions, pedagogical mismatches, genetic predispositions, and even prenatal exposure to toxins. In this study, we adopt a broad definition that includes clinically defined difficulties (with known ICD codes) as well as difficulties reported in the school environment.

There are several ways to address learning difficulty, from alleviating it or stopping it. For example, active methodologies can be used to motivate students, which may involve new technologies and tools such as computers and robots. One such active learning approach is Problem-Based Learning (PBL), which engages students in solving real-world problems using engineering methods1. Together with this technique, the use of robots in education has increasingly gained attention2,3,4,5.

Educational robotics has gained increasing prominence in recent years due to its transformative potential in traditional educational practices and its promotion of active, practical, and multidisciplinary learning environments5,6. More than just another tool, it has been increasingly considered a promising strategy to address some of the most pressing issues in education, such as low academic performance, demotivation, and challenges in inclusive practices involving students with learning difficulties7. However, although advances in educational technologies have broadened the possibilities for learning, students with learning difficulties remain among the most disadvantaged groups, with limited access to these innovative pedagogical activities8.

Generally, Educational Robotics is used with two main purposes (or approaches)6: (1) as a subject in itself—teaching concepts in robotics such as mechanics, electronics, and programming, among others; or (2) as a teaching support tool—to support the teaching of other subjects. The first approach is commonly seen in industry training and in the preparation of students for competitions such as the Brazilian Robotics Olympiad (OBR)9 or the First Lego League (FLL)10. This approach satisfies the clear objectives of the Industry 4.0 paradigm11,12 and is often directed at students with higher academic performance prior. However, it is the second approach, robotics as a learning aid tool, that has sparked interest among educators in a wide range of fields committed to inclusive education and true access to knowledge6.

The second approach—robotics as a learning aid tool—has become especially relevant to education, particularly in the context of inclusive and interdisciplinary teaching practices5. This approach allows robotics to be used in different subjects, including languages, science, history, geography, and physical education, among others. However, its effective implementation requires teachers of these subjects to acquire at least basic knowledge of robotics and its pedagogical potential13.

In Brazil, initiatives to institutionalize educational robotics in public education systems are advancing, supported by public policies and various partnerships. An example is Law No. 0588 of the Natal City Council, enacted on June 26, 201914, which integrates robotics into elementary education through activities coordinated by the Municipal Education Secretariat (SME). Collaborations with institutions such as the Federal University of Rio Grande do Norte (UFRN) and the Catholic University of San Pablo (UCSP) in Peru have been essential. The Municipal Reference Center for Education (CEMURE), in partnership with these universities, has offered continuing education in educational robotics, contributing to the training of qualified teachers capable of applying robotics in the school environment6,15,16.

Even in schools where educational robotics activities are implemented, the participation of students with learning difficulties remains remarkably low17. In our review of the literature, we identified few studies that specifically address this issue, and, among them, the presence or participation of these students in robotics activities is often minimal. Conchinha17, for example, observed that despite the proposal of robotic workshops, students with learning difficulties did not show an interest in participating. However, when these students were motivated by the multidisciplinary team, they became fascinated by robotics. These findings in the literature allowed us to identify two main research gaps: (1) the limited number of studies focusing on the inclusion of students with learning difficulties in educational robotics; and (2) in the existing literature, the lack of concrete strategies or evidence of meaningful participation of these students in such activities.

These gaps substantiate the need for targeted investigations that examine not only access but also meaningful participation and educational outcomes for students with learning difficulties. Rather than focusing exclusively on robotics as a technological resource, this study approaches it as a mediator for the development of skills and competencies in students who often face barriers in traditional pedagogical contexts. Educational robotics, when used inclusively, can promote engagement, autonomy, and collaborative learning, helping to reduce exclusion and demotivation in the school environment18,19.

The central research question guiding this study is: How can educational robotics be effectively integrated into pedagogical practices to support the learning process of students with learning difficulties? Based on this, the study seeks to identify challenges, limitations, and opportunities in the implementation of robotics activities, particularly in the context of students with learning difficulties.

To do this, we adopted a qualitative and quantitative approach. First, a systematic review of the literature was conducted to identify the gaps between robotics and inclusive education. Next, a field study was conducted in public schools in Natal, RN, Brazil, which are part of the robotic activities program in partnership with UFRN and UCSP. The data obtained involved the application of questionnaires to students, teachers, and trainees participating in robotics activities.

We emphasize that the preliminary scientific contributions of this work were disseminated to the academic community through two conference articles published in the Workshop on Educational Robotics (WRE). One in 2022, where we initially mapped the state of the art using a systematic review approach, with the first article on this topic20, and another in 202321, which laid the foundations for the present study and was selected among the best articles from WRE 2023 to be submitted in this expanded version. Based on these contributions, this article expands the research by detailing the case study on the application of educational robotics in public schools in the municipality of Natal.

In addition to this introduction, the rest of the article is structured in the following sections: “Background” presents the conceptual framework of educational robotics and learning difficulties. “Related works” discusses related works that explore robotics in inclusive education. “Methodological procedures” describes the methodological procedures adopted in the study. “Results and discussion” presents and discusses the results obtained in the field research. Finally, “Conclusion” describes the main conclusions, limitations, and future directions of the research.

Background

In addition to the fact that there are a number of interesting studies on topics within the field of educational robotics, our main proposal is to understand and arrive at a model of how these new forms of learning (with robots) should be used with students who have some kind of learning difficulty or disorder, and to try to understand how much this approach actually helps with learning. In order to better understand our study, we present in the next the basic assumptions and ideas of educational robotics and also conceptually present what learning difficulty is, so that the reader can better understand the context of our work.

Educational robotics

The use of educational robotics began with Seymour Papert22, who introduced the constructionist approach, a pioneer in this technology22,23. This paradigm uses practical learning strategies in order to provide knowledge, by working with the construction and programming of robotic devices. When Papert first asserted that children would use computers, especially in the context of education, it was met with skepticism and sometimes even amusement. This was primarily because during the time when Papert introduced these ideas in the late 1960s and early 1970s, computers were large, expensive machines primarily used for scientific and business purposes. The idea of children, who were not seen as capable users of such advanced technology, engaging with computers seemed far-fetched to many. Today, this is a reality, not only for computers, but also for robots.

Corroborating Papert’s aims, a significant group of researchers emerged from the 2000s, influencing the implementation of robotics in schools. There are researchers investigating mechanics, electronics and computing, even from industry, who have contributed to making the current fascinating and innovative reality of robotics possible, inspiring generations of teachers and students4,24,25,26,27.

Thus, educational robotics was born as a somewhat modern field of investigation and has spread methodologies, arising a certain curiosity in the school community2,3. It can be used from the early stages of education up to higher education levels28 and has been consistently adopted as a new methodology in the teaching and learning process throughout the world. Basically, the most widely used model consists of encouraging students to program and follow the behavior of robotic prototypes4,26, and even use their own programming languages, whether graphical or textual. Thus, most teachers and monitors (trainees) are using the learning methodology known as Problem Based Learning (PBL)1.

However, there are various challenges that need to be addressed for different applications of educational robotics. Schools and industries adopt it in short-term courses in practice, often without considering the continuity it demands. Despite this, it represents a valuable pedagogical resource for both learning and training, with the potential to be permanently incorporated into the school educational program29. Furthermore, recent advances have been made in educational robotics, particularly in terms of hardware platforms and software environments2,30,31.

Although, as already mentioned, there is some reluctance, especially on the part of teachers, to effectively introduce educational robotics as part of their educational practices, even if only once a week32. The main reason for this, in most cases, is the lack of knowledge of the benefits that this technology can bring to the learning of any content at school. Despite this, some teachers are interested in adopting it33.

In addition, educational robotics is known to be an interesting methodology for supporting teaching and learning practices. Specifically, it is an attractive tool for students, helping them to get involved in Science, Technology, Engineering, Arts, and Mathematics (STEAM) careers4,26. It is known to be a powerful tool that enables the development of material to engage these five disciplines. In fact, it has been argued and proven that it can achieve even more than traditional approaches5.

In this sense, in order to begin to tackle these issues, some work has been carried out with completely blind students18 and also with students who have some degree of Autistic Spectrum Disorder34. This work takes a step in the same direction, but we are dealing with students with learning difficulties, general or specific.

Learning difficulties

Learning difficulties have been a subject of attention across various fields of knowledge, particularly due to their impact on students’ academic trajectories and their underlying causes. In this study, we adopt a broad approach to these difficulties, expanding on our previous work20, in which specific cases—with or without clinical diagnosis—were analyzed based on the application of educational robotics as a pedagogical resource. Now, we seek to deepen this discussion and investigate whether robotics can assume a role that goes beyond a simple tool, becoming a transformative educational strategy capable of promoting the development of skills and competencies in students.

The term learning difficulty was first proposed by Kirk in 1968, with a definition introduced in 1969. It is worth noting that he often refers to the term “disabilities” or “difficulties” for the same issue. It refers to difficulties in one or more of the basic psychological processes involved in understanding or using spoken or written language35. These difficulties manifest themselves in various ways, such as problems with hearing, thinking, speaking, reading, writing, spelling, or performing mathematical calculations. The definition covers conditions such as dyslexia, minimal brain dysfunction, perception disorders, and aphasia. However, it explicitly excludes learning problems that are primarily the result of visual, hearing, or motor impairments, intellectual difficulties, emotional disorders, or sociocultural35.

On the other hand, Fonseca36 considers that learning difficulties are characterized by a significant discrepancy between an individual’s cognitive potential and their actual academic performance, occurring in students who do not have intellectual, sensory, or emotional impairments. These difficulties can affect basic areas such as reading, writing, mathematics, or areas such as spatial orientation, motor coordination, or social interaction. The author emphasizes that these challenges are neurological in nature and are not due to a lack of motivation or effort.

In the same direction, Smith and Strick8 point out that many students with learning difficulties are often misunderstood by educators and families, and are frequently mislabeled as having low intelligence or being lazy. In reality, they face neurological or cognitive challenges that affect their ability to process, retain and communicate information effectively.

Due to the diverse nature of learning difficulties, many authors emphasize the complexity of providing a concise and universally accepted definition8,35. Consequently, the concept has been explored from various disciplinary perspectives. From a medical point of view, learning difficulties (or disabilities) are often associated with neurobiological factors. In contrast, behavioral and psychological perspectives tend to focus on aspects such as perception, memory, attention, and cognitive processing. Kirk and Gallagher35 propose three essential criteria for diagnosing learning difficulties:

  • Discrepancy criterion: a significant gap in cognitive or academic development, often without a clear cause. For example, a student may perform well in math, but may not acquire basic reading skills.

  • Exclusion criterion: the difficulties cannot be primarily explained by other conditions, such as intellectual difficulties, sensory impairments, or emotional disorders.

  • Special education criterion: the student requires specialized teaching methods to overcome learning barriers.

To better synthesize and understand this diversity of viewpoints and their key contributions, Table 1 outlines the main approaches to learning difficulties, emphasizing their conceptual focus, key authors, and relevance to educational practice.

Table 1 Main approaches to learning difficulties.
Full size table

Therefore, in light of the diverse perspectives on learning difficulties, researchers have sought to understand both the causes and the factors that hinder students’ ability to learn, including brain dysfunction, genetics, environmental and nutritional factors, as well as biochemical factors, considering etiology an important aspect in preventing or improving these difficulties8,35. However, instead of exclusively focusing on the origin, other authors argue that it is crucial to understand the impact of these difficulties on the student’s perception of the world and to find the most effective strategies to contribute to their educational development8. This idea is somewhat used in the current work. We try to apply educational robotics tools to minimize or alleviate learning difficulties.

Related works

In this section, we present and discuss the main works found in the literature relevant to our topic, including articles as well as theses and dissertations. Here we will go into more detail about the strictly related works that were used as references, highlighting their main contributions, limitations, and differences from our work. Indirectly related works, which have some relevance to the topic but are not directly applicable, will also be discussed.

Peer-reviewed articles on educational robotics and learning difficulties

This subsection synthesizes peer-reviewed scholarly articles (articles in events proceedings or journal papers) on Educational Robotics and learning difficulties. The goal here is to consolidate validated empirical findings and delimit strictly related works that speak directly to our research question. We used a systematic review approach37, which provides a smaller number of works that are strictly related to a topic through a more refined search. This approach is suggested for a more detailed and in-depth analysis38.

Our initial search was carried out using the CAFE platform39 just to enter the research bases. It provides access to more than 120 databases in which a registered user can perform a search. The set of databases available includes the most widely used, such as IEEE Xplore, Springer Link, ACM Digital Library, Web of Science (Core Collection), Scopus (Elsevier) and Pubmed (NCBI Literature Resources). Our general search expression simply concatenates the keywords, depending on the database used, such as: (Educational Robotics AND Learning Difficulty) OR (Educational Robotics AND Learning Disorder) OR (Educational Robotics AND Autism Spectrum Disorder) OR (Educational Robotics AND ADHD).

Entering these terms into the various databases, we carried out the searches and, after eliminating duplicates provided by each database, we obtained 50 scientific papers, which were detected from the six databases mentioned above. It is worth noting that in the search engines and in our searches we used keywords that represent any difficulties, whether they stem from social or particular problems of the students or due to causes with well-defined medical codes, such as ASD, ADHD, dyslexia, among others. We also carried out a search including other sources of consultation, in order to find all the works we considered relevant to our topic.

After an initial screening, reading the titles and, in some cases, the abstracts of all the documents, we identified 28 articles that were incorporated into our state of the art. Of these 28 articles, only two were strictly related to our theme, demonstrating the relevance and novelty of our work. The first article17 directly addresses the topic of learning difficulties and educational robotics, and the second article18 addresses the terms indirectly. As the second work is also published in the form of a thesis, it will only be discussed in the next section, which deals with theses and dissertations found on the subject in Brazil. Thus, we present the state of the art, using a systematic literature review approach, with closely related works shown in Table 2.

Table 2 Comparison of related studies.
Full size table

The article of Conchinha17 considers any learning difficulties. It reports on activities in an educational robotics project, in which, among the universe of students, three ninth graders took part who had specific learning difficulties, but without a defined medical report due to lack of resources, as the authors point out in the article.

The authors report that the participating students had low academic performance. One had a lack of interest in regular classes because he could not see any prospects for his future in school activities. Another student had difficulty in reading, interpreting, and understanding what was being read and also lacked attention, with very frequent distraction in class. One student also had limited academic difficulties with logical reasoning, which has impacted his performance in mathematics.

With the proposal of activities with educational robotics, the students were reluctant to initially proceed with them. So, a starting dialog was proposed where the students were asked about their opinion on educational robotics, since the school already had a robotics project at times other than regular classes, but none of them had signed up for the activities because they were not interested, as they reported in their answer. The students’ expectations of the proposed robotics activities were also provoked, and the Lego kits to be used were presented, from which point the students began to explore the robotics parts and prototypes.

As the activities developed with the students chosen because they had specific learning difficulties, the authors report that one of the students was shy to begin with and another withdrew, observing only the other classmates. Later, this student took the initiative and began to select pieces and give instructions to others, and in the end they all took an active part in the activity.

The students then realized the importance of activities, especially playful and project-based ones, as they motivated and challenged them, as reported by the authors17. In addition to working on a range of skills and competencies, through reading and interpreting instructions, assembling and programming robotic prototypes, team work and logical reasoning.

Theses and dissertations on educational robotics and learning difficulties

This subsection aims to delve even deeper into the our state-of-the-art. We carried out an advanced search for theses and dissertations using the keywords Educational Robotics and Learning Difficulties. We used a publicly available repository, the Brazilian Digital Library of Theses and Dissertations (BDTD), which allows access and visibility to theses and dissertations40. This specific database includes 139 institutions and contains 898,982 documents, including 656,162 dissertations and 242,820 theses. Unlike journal/conference articles, these documents often provide extended methodological detail, contextual descriptions, and exploratory or emergent evidence. We use this corpus to complement the peer-reviewed articles in “Peer-reviewed articles on educational robotics and learning difficulties”, identifying related research and methodological insights that inform our study design.

Our initial search resulted in 31 works from the BDTD database. By observing in more detail these 31 works, three were published twice, which we removed, leaving 28 works. When analyzing each of these 28 works, only one did not use robotics in its theme, and it was the last to appear in the search. In this way, our search for the keywords Educational Robotics and Learning Difficulties found a total of 27 works, 22 of which were dissertations and 5 theses.

We found previous works that use educational robotics to teach curricular subjects such as physics, mathematics, and English from primary to secondary school. In addition, we identified works focusing on educational robotics for teaching programming logic and computer vision to higher education students.

In line directly with our research theme, we identified only two among the theses and dissertations that contemplate our context of educational robotics and learning difficulties, which are shown in Table 3.

Table 3 Comparison of theses strictly related to this study.
Full size table

The first thesis41 presents the experience of students who are identified as having learning difficulties by the school in classroom instruction activities and robotics workshops, and how these activities could contribute to creative practices and the cultural development of these students.

After a thorough analysis of the thesis mentioned above41, we identified that the research participants were five students considered to have learning difficulties by their teachers. It was also identified that the robotics monitors’ view of the students was different from that of the teachers. This difference was reflected in the different ways in which students participated in the classroom and in robotics workshops. This difference can be seen in the results presented in the thesis41, which concluded that many of the so-called learning difficulties are social and school difficulties and not biological or psychological.

Then, this research addressed extrinsic conditions for student learning35. And external conditions do not consider students with learning difficulties, because they are not causes inherent to the student, but external causes, such as the teachers’ view of the students and the monitors’ view, which differ from one another41.

In the second thesis, which we also consider to be close to ours, educational robotics is applied to visually impaired children. The purpose of the work is to provide access to educational robotic activities for students with visual impairment or low vision, using CardBot 2.019. The authors have developed a technique for applying educational robotics based on cards and a programming interface, which the child can use to program and pass on commands to the robot.

With this, the child (with visual impairment or low vision) can learn robotics through inclusion and the appropriate use of robotics activities for their reality. An important contribution of the work is that the teacher can create new cards and record the marks of the respective actions19. This allows the teacher to add new actions for the robot or even create a new language that respects the student’s level of knowledge.

The proposal in the second thesis19 was validated with experimental lessons for students with and without visual impairments and different age groups. So, although the work in question does not look at general cases of learning difficulties, a specific instance (children with visual impairment or low vision) is addressed.

Taken together, “Peer-reviewed articles on educational robotics and learning difficulties” and “Theses and dissertations on educational robotics and learning difficulties” offer complementary views: peer-reviewed articles provide consolidated empirical evidence, whereas theses and dissertations supply methodological depth and emergent results. We now position our work in light of these two lenses.

Positioning our work

We identified several aspects in the state of the art that inform and contribute to the inclusion of students with learning difficulties in educational robotics. Our systematic mapping20 found few strictly related peer-reviewed articles and only two closely aligned theses/dissertations, highlighting that this topic remains largely unexplored.

As introduced in “Peer-reviewed articles on educational robotics and learning difficulties”, Conchinha17 reports an extracurricular robotics experience with LEGO kits involving three ninth-grade students who exhibited low academic performance, reading/comprehension difficulties, attention problems, and limited logical reasoning, all without formal medical diagnosis due to resource constraints. Initial reluctance gave way to progressive engagement after dialog and hands-on exploration, with perceived gains in motivation, teamwork, logical reasoning, and problem solving. However, the extracurricular setting, very small sample, absence of standardized instruments, and lack of longitudinal follow-up limit generalizability and obscure whether perceived gains translate into sustained classroom outcomes or curricular integration—precisely the gap our triangulated design seeks to address.

In “Theses and dissertations on educational robotics and learning difficulties”, Villaça (2022)41 analyzes students identified by the school as having learning difficulties in both classroom and workshop contexts, revealing divergences between teachers’ and monitors’ perceptions and underscoring social/school determinants—external to the learner—as contributors to “difficulties”35. While theoretically valuable, the evidence rests on a small sample and heterogeneous informant perspectives, restricting external validity and motivating designs with standardized measures and multiple stakeholders. Also in “Theses and dissertations on educational robotics and learning difficulties”, Pitta (2017)19 proposes CardBot 2.0 and a tactile programming interface to make robotics accessible to children with visual impairment/low vision, advancing assistive design and teacher adaptability (e.g., creation of new tactile cards) and validating the approach through experimental lessons. As a condition-specific intervention, it does not address broader learning difficulties or cross-subject outcomes.

Our work situates educational robotics within public-school settings and addresses a broader field of learning difficulties through a mixed-methods design that triangulates perspectives from students, teachers, and interns. We adopt an operational definition that encompasses both clinically defined conditions (per the ICD) and context-related barriers commonly reported in schools, and we look beyond mere access to robotics to probe evidence of meaningful participation and learning outcomes. Rather than treating robotics solely as a technological resource, we position it as a pedagogical mediator aligned with inclusive education and consider complementary modalities such as PBL and simulation environments42. In doing so, we extend prior work—moving past single-case demonstrations and condition-specific assistive designs—toward a school-embedded, inclusion-oriented perspective with clearer implications for external validity and curricular integration.

Research on humanoid robots has also examined contributions to learning for children with Autism Spectrum Disorder (ASD)43, noting both potential benefits and instances of avoidance; as a practical alternative, non-humanoid mobile robots have been explored. While adjacent to our focus, these findings support the broader premise that tailored robotic mediation can foster engagement under diverse constraints20.

In summary, prior studies either examine small extracurricular cohorts without standardized measures (e.g.,17), address a specific disability via assistive design (e.g.,19), or highlight social/school determinants in limited samples (e.g.,41). We respond to these gaps with a school-based, mixed-methods investigation centered on meaningful participation and learning outcomes for students with learning difficulties.

Methodological procedures

We first carried out a literature review to acquire theory on the matters of Educational Robotics and Learning Difficulties. Next, we provide a systematic review to identify whether there are strictly related works that address our theme. The list of related works was then completed with theses and dissertations on both topics found in repositories, in order to define the state of the art and best establish the theoretical framework. Finally, to this end, we apply questionnaires to students, teachers, and trainees (being a form of triangulation approach) from public schools with educational robotics activities in the municipality of Natal, RN, Brazil. Thus, as a research methodology, we adopted a qualitative-quantitative (quali-quanti) approach44, with the support of questionnaires as data collection instruments. In the quali-quanti approach, qualitative but also quantitative aspects are considered in order to infer and/or verify whether a given thesis or hypothesis is valid. In education specifically, quali-quanti research makes it possible to describe the phenomena observed by the researcher and to substantiate the views obtained through evidence45. Although quantitative methods are prevalent in the exact sciences, qualitative analysis plays an important role, particularly in understanding complex phenomena that cannot be easily quantified. Our approach involves identifying and analyzing patterns (themes) within qualitative data to highlight significant themes, allowing a deeper assessment of the underlying issues and insights that might not be evident through quantitative analysis alone46. Hence, as mentioned above, in this work, we used this mixed approach based on questionnaires that were administered to teachers, monitors, and students who participated in educational robotic activities.

Furthermore, our study was conducted in accordance with the guidelines and regulations of the Ethics Committee of the Federal University of Rio Grande do Norte (UFRN). The study has the Certificate of Presentation and Ethical Assessment number 78048724.1.0000.5292, generated after its approval for execution. We emphasize that the consents were collected from the research institutions, as well as the informed consent forms and the authorizations for audiovisual recordings, of all research participants and their respective legal guardians. Considering that every research process can generate some risks, we emphasize that: the confidentiality and privacy of the participants were ensured; the non-violation and integrity of the archived documents; the interviews were conducted in appropriate and reserved places to guarantee the right and freedom of expression; the research data are archived in digital and analog format, without risk of access by third parties or data leaks, for a period of five years, being for exclusive use of the research for scientific investigation and dissemination, and will be properly discarded after the aforementioned period.

Questionnaire preparation

We planned and developed the questionnaire with contributions from Pedagogy professors at the Education Center of the Federal University of Rio Grande do Norte (UFRN), together with members of our Robotics in Education research group. Our current research included six public schools in the city of Natal, RN. Robotic kits were given to these schools and the professors have participated in a pilot project to implement educational robotics activities. This initiative is aligned with the public policy established by the municipal council of Natal, RN, Brazil14.

We devised similar structure questions for students, teachers, and trainees, with specific adaptations to each research topic. A form was made available on the Web for each type of respondent subject, which they answered the questions online. The most relevant questions for students are such as:

  • To what extent do you believe robotics has enhanced your learning (on a range of 1 to 5)?

  • Have you faced any problems during in the robotics activities? Please, answer yes or no? If you answered yes, what were these problems about?

    ( ) The assembly the robot

    ( ) The robot programming

    ( ) Other

  • Please, write with your own words how do you think the use of the robot helped you to learn better the class content, if so? (free text)

  • What is most attractive to you when learning robotics? (free text)

The questions for teachers and trainees are presented below, with basically the same structure as those for students. However, it is clear that our intention was to dissociate the questions for students presented above in order to identify the perception of teachers and trainees about the contributions of robotics to students’ learning difficulties.

  • Do you have students with learning difficulties in your class? Which ones?

    ( ) ADHD

    ( ) TEA

    ( ) Dyslexia

    ( ) Reading/Writing

    ( ) Others

  • Quantify the motivation of students with learning difficulties to study using robotics (on a scale from 0 to 10).

  • What are the difficulties encountered in implementing robotics at school?

    ( ) Lack of knowledge about robotics

    ( ) Lack of infrastructure at school (lab, kits)

    ( ) Lack of classroom conditions (materials, kits)

    ( ) Lack of time for planning and preparing activities

    ( ) Lack of teaching materials

    ( ) Lack of support from managers

    ( ) Others

  • Do you think that robotics should be introduced at school? If so, for what purpose?

Once again, it is important to emphasize that the main objective and premise of this research is to identify whether educational robotics, used as a pedagogical resource, contributes to students who have learning difficulties or disorders. And if this occurs with any student or group of students, what impacts can be observed on their learning?

When administered the questionnaires (see below), we realized that there were in fact several students with general learning problems (disorder or difficulty), but who are not actively participating in robotics activities. This was initially recognized in strictly related works, but also in our preliminary visits to the schools and also at the time of administering the questionnaires, when we found that a small number of students needed help from our team to answer the questionnaires, due to the (enormous) difficulty in reading/writing that some of them had.

Furthermore, is also recognized in strictly related work and preliminary visits to schools, we noticed a lack of initial interest on the part of the students in the possibility of using educational robotics to support the learning process. However, when they were introduced to the robotics kits and what they could do for their lives, in terms of developing skills and competencies, as well as providing more fun and entertaining learning, the students really began to take an interest in this tool.

Questionnaire application

The questionnaire was administered in six schools in Natal, RN that joined the project in partnership with the Municipal Education Secretary and UFRN, which consolidated the public policy of implementing educational robotics as a complementary activity. Some of the teachers in these institutions participated in the FOCORE course and accepted the proposal to include educational robotics in their teaching practices6,15,16.

The average age of the participating teachers is 47.6 years, with a standard deviation (SD) of 12.1 years, demonstrating significant age diversity among the research subjects. This variation contributes to a broader understanding of perceptions about the use of educational robotics, considering different professional experiences and teaching generations. The subjects that each of them said they teach are: one in Arts, one in Physical Education, one in Geography, and two teachers teach Mathematics. In addition, one of the teachers reported that he is responsible for activities in the Computer Lab, which caters for primary school students. The other teachers work with robotics in their subjects with elementary-school students.

All teachers have higher education qualifications. Specifically, the following qualifications are held: Bachelor’s Degree in Arts/Music (1 teacher), Master’s Degree in Geography (1 teacher), Master’s Degree in Physical Education (1 teacher), Bachelor’s Degree in Mathematics (1 teacher), Full Bachelor’s Degree (Teacher Training Course for 1st to 4th Grade Elementary School) (1 teacher), and Postgraduate Degree in Mathematics (1 teacher). Teachers who participated in our research did not report additional certification. Of the six teachers interviewed, four are women and two are men.

These teachers were chosen to make up our group of teachers interviewed because they had already taken the Continuing Education in Educational Robotics (FOCORE) course offered by CEMURE or because they were or had already had the experience of applying educational robotics in their teaching practices. However, some of the teachers participating in our research reported that they had not experienced the FOCORE course but saw the proposal to work with the help of robotics as interesting and joined the project, even without understanding the basic concepts to start their journey.

This fact corroborates the importance of robotics as a way to stimulate the interest of educators in various fields35. Another important aspect of the teachers’ adherence, even if they have not had first contact with robotics, was the presence, motivation, and fascination of the trainees in the schools, who encouraged these teachers to take a closer look at robotics in the educational context.

Therefore, there were six trainees who participated in our investigation and who also answered the questionnaire, the same number as schools that implemented educational robotics as a complementary activity. And, like the teachers participating in the FOCORE course, the trainees were also expected to participate in a training course to work alongside the teachers on the robotics projects. However, some of them did not participate, which led to different action methodologies on their part.

We emphasize that this had no direct impact on the results, as they were all motivated and were all undergraduate students in the field of Science and Technology, which implies that they had the necessary knowledge of the basic principles of the problem-solving learning methodology used in our educational robotics workshops.

The students who participated in the study were between 11 and 16 years old, with a mean age of 12.3 years and a standard deviation of 1.3 years. Thirty-three students were identified, 21 boys and 12 girls. All were enrolled in municipal public elementary schools in Natal, RN, in grades 5 through 9. These students had already participated or were participating in educational robotic activities in their schools and were invited to voluntarily complete the questionnaire. It is important to note that the selection sought to identify students who had presented learning difficulties, regardless of whether they had a medical diagnosis or classification.

In all schools, activities were organized according to the complementary activities calendar and typically involved hands-on tasks (assembly and programming), teamwork, and problem-solving (with the help of trainees). According to student responses, the most frequently reported activities involved assembling and programming robots with LEGO EV3, while difficulties centered around these same tasks (see Figs. 3 and 4). The pace and duration of the activities were those adopted by each school’s robotics curriculum (e.g., weekly 60–90-min sessions throughout the school year). We report this to contextualize the learning environment in which participants experienced educational robotics activities, rather than prescribing a single protocol for all schools.

All students are from municipal public schools in Natal, Rio Grande do Norte, and are involved in an educational robotics project, in accordance with Law No. 0588/2019. The distribution of students by grade is as follows: 3 5th graders, 19 6th graders, 3 7th graders, 5 8th graders, and 3 9th graders.

Some of the students had a defined medical report, but most of them did not have a report proving that they had learning difficulties, whether in the development of spoken or written language, arithmetic, or even a medically defined disorder35. However, during the course of the investigation, these students were identified, regardless of the presence of a medical report.

In this pilot project, we have not performed a clinical diagnosis to identify students with learning disabilities. Instead, we adopted a broad definition of learning difficulties aligned with our study objectives and the school context. Using this approach, students could be included if they had a formal report (with ICD) or if their teachers and/or trainees consistently reported persistent learning challenges (e.g., reading/writing, arithmetic, attention/behavior) observed in the classroom and in supplementary robotics activities. Identification, therefore, was based on triangulation of sources: teacher reports (questionnaire items on the presence and type of difficulties), trainee reports (parallel items on observed difficulties), and student self-reports, using structured questionnaires developed with educational experts and approved by the institutional ethics committee. This procedure aimed to identify educationally relevant difficulties for inclusion in a context where not all students had undergone prior clinical evaluation in the school system.

It is important to note that we are carrying out this research as part of a pilot plan to introduce educational robotics activities in public schools in the municipality of Natal, RN, following the enactment of law Nr. 0588, of June 26, 2019, which is being implemented through various partnerships and which will take up to eight years to gradually introduce educational robotics throughout the municipal education network.

Therefore, to administer the questionnaires, we had the collaboration of 33 students, six teachers, plus six trainees that are university students working at schools on the robotics project. This instrument was chosen because it allows the researcher to capture the interests, values, and motivations of a considerable number of subjects47.

With regard to data analysis, the learning data were provided by teachers and are reliable. Other data, such as student performance, are verified by the trainees, who are undergraduate and graduate students working under our supervision, so they are also reliable. In addition, we argue that all of the subjects are part of a pilot plan worked out in order to introduce educational robotics as a complementary activity in the municipality of Natal, RN, Brazil, which is the test field of our research. So, due to the fact that law No. 0588, of June 26, is still in the process of being complied with, since 2019, and each year should reach a percentage of the schools in the municipality, we do not have a huge amount of teachers and trainees working with robotics to compose our research. So, this explains why we used the adopted quali-quant methodology, which is suggested to be applied in such cases45.

To synthesize the sequence of activities carried out during the research, Fig. 1 illustrates the main stages of the methodological procedure, from the thematic study to the analysis of the collected data.

Fig. 1
Fig. 1
Full size image

Stages of the research methodological procedure.

Results and discussion

After having collected data using the procedures described in “Methodological procedures”, in this section we present the results that were obtained from the analysis of the various questionnaires that were applied to students, teachers and trainees, dealing with the possible contributions of educational robotics for students with learning difficulties or disorders.

Student questionnaire

We begin by showing the results of the student questionnaire with questions about the influence of robotics on their own learning processes. As for the “influence of robotics”, we explained to them that we wanted to know how much they believed robotics could help, quantifying whether or not they would like robotics to be present in their classes in general.

Fig. 2
Fig. 2
Full size image

Robotics’ contribution to student learning.

The results indicate that most of the students reported that robotics contributed positively to their learning (Fig. 2). Specifically, eight students assigned a score of 3, nine assigned a score of 4, and 12 rated it with the maximum score of 5. The score of 2 or less was given only by four students. The mean score was 3.8, with a standard deviation of 1.22. In general, 87.9% of the respondents recognized robotics as beneficial in enhancing their learning process.

We have identified positive results that are corroborated by the students’ responses in free text when asked about their interest in the learning process when the teacher uses robotics to mediate some content. When considering this result, it is impressive to see how students understand the benefits of robotics in their own learning. We obtained several answers, including some very interesting ones that have been compiled and are presented in Table 4.

Table 4 Answers from students when the teacher uses robotics.
Full size table

Later, the students were asked if they have had any difficulties with the robotics activities. About 57,6% of them answered yes to this question and 42,4% answered no, as can be seen in Fig. 3. In other words, 14 students answered that they had no difficulties and 19 students answered yes to this question.

Fig. 3
Fig. 3
Full size image

Percentage of students with and without difficulties in robotics activities.

For those who answered yes, it was also necessary to further answer what these difficulties were related to robotics activities (Fig. 4). We remark that this is a multiple-choice question for students who had encountered some difficulty. In this question, students could select more than one alternative or even express their answer in free text.

Fig. 4
Fig. 4
Full size image

Types of difficulties encountered in robotics activities.

Fourteen of them reported difficulties in assembling the robot and thirteen encountered difficulties in programming the robot (see Fig. 4). Two students pointed out specific difficulties and reported this in free text. One of them reported that he found it difficult at first to assemble the robot without the step-by-step assembly instructions. He considered it more important to learn with the step-by-step instructions. Another student pointed out difficulties in Arduino programming. Only one student selected yes, without expressing any difficulty. The students were also asked what they liked more in learning robotics and we obtained several interesting answers that can be seen in Table 5.

Table 5 Students’ answers about learning with robotics.
Full size table

We noticed in the students’ answers about what they most liked to learn in robotics (Table 5), among various aspects, an interest in STEAM4,26 areas. It is clear that robotics has motivated and encouraged students to learn about assembling and programming robots, as well as developing teamwork, creativity and the desire for new challenges.

It can be noticed that according to the enactment of law No. 0588, of June, 26th, 2019, after it comes into force, each year there is a percentage of implementation reach in schools. According to the law, in the first year that it came into force, robotics as a complementary activity should be offered to 10% of elementary school students in public schools in Natal, RN and so on each year until it reaches all students in the municipality14. As a result, it is possible to understand that the law declared in 2019 is still being systematically implemented and that a small sample still made up our research.

Teacher questionnaire

The initial question directed at teachers investigated whether they had students with learning difficulties or disorders in their classes, and if so, what types. Approximately 66% reported having students with these challenges, most often difficulties in reading and writing, which may be associated with the pandemic period.

Fig. 5
Fig. 5
Full size image

Learning difficulties.

Teachers also identified specific conditions, including Attention Deficit Hyperactivity Disorder (ADHD), Autism Spectrum Disorder (ASD), and dyslexia. One teacher, under the “Other” option, noted the presence of students with diverse difficulties across grades 1 to 5, but without specifying the exact disorders (see Fig. 5).

Fig. 6
Fig. 6
Full size image

Motivating students with study difficulties through robotics.

Teachers were also asked to evaluate, on a 0–10 scale, the motivation of students with learning difficulties after engaging with robotics. The responses revealed considerable variation: two teachers rated motivation as 10, one as 9, one as 2, and two as 5 (Fig. 6). These results suggest divergent perceptions—some teachers recognize robotics as a facilitator of learning for students with difficulties, while others remain skeptical.

Such divergence may be linked to the fact that robotics is still in the early stages of systematic implementation in the municipality, and its benefits for student learning are not yet fully understood14.

Finally, when asked about challenges in implementing robotics in schools, teachers mentioned not only predefined alternatives but also provided open-ended responses, highlighting specific difficulties in more detail (see Fig. 7).

Fig. 7
Fig. 7
Full size image

Difficulties in implementing robotics.

The teachers’ free-text responses on the difficulties they encountered in implementation were, respectively:

  • “Lack of knowledge of students and resistance to new types of learning”;

  • “The weekly service in the computer room provides different pedagogical experiences for 11 elementary school classes. It’s difficult to manage robotics in a more meaningful way”;

  • “Reading and writing problems”; and

  • “Lack of interest from teachers”.

We understand how many challenges teachers face as a result of students’ learning difficulties. However, we can see that there is some interest and also motivation from the teachers in working with robotics. Especially when they were asked if they were interested in robotics being implemented in schools, and, unanimously, all of them said yes (Fig. 8), despite the challenges encountered (Fig. 7). This was not surprising, as most of these teachers had already been previously motivated as they have taken courses on robotics as a supporting tool at the CEMURE (Municipal Center for Reference Education), in the municipality of Natal, RN6,38.

Fig. 8
Fig. 8
Full size image

Interest in implementing robotics in school.

In addition, for the question about the implementation of robotics in their schools, teachers were also asked about the purpose of the implementation, and the answers are shown in Fig. 9. It should be noted that more than one option could be selected for this alternative. The six teachers have answered the survey by selecting the option that uses it as a teaching tool in their subjects. Four of them also chose the option to help students with learning difficulties and two teachers chose the option to prepare for competitions, such as the OBR and the FLL.

Fig. 9
Fig. 9
Full size image

Purpose of implementing robotics at school.

We should point out that after robotics was introduced in schools, one of the six schools has participated in the FLL robotics tournament, in its regional phase48. When examining this work, we realized that a student taking part in the tournament had a medically defined coding disorder, in particular ADHD, and we saw how much his limitations overlapped with his potential in robotics activities, which required logical thinking and teamwork. This same school has also participated in the OBR, in its regional phase, receiving the title of Best Public School49. We realized how much robotics has been observed to ignite and promote effective gains for teachers and trainees, as well as for students, including those with learning difficulties or disorders, within the scope of this pilot study. We note that additional studies with larger and more diverse samples are necessary to confirm and expand these findings.

Trainee questionnaire

We also began by asking the trainees if there were any students in their classes with any learning difficulties or disorders. Four trainees pointed out the presence of students with ADHD, two selected the presence of students with ASD and also with reading/writing difficulties and only one trainee indicated students with dyslexia present in their classes (see Fig. 10).

Fig. 10
Fig. 10
Full size image

Learning difficulties or disorders.

We also asked the trainees how much robotics improved the learning process of students with learning difficulties or disorders. We noticed that only two trainees selected the value 5, which is the maximum grade for this alternative, three of them selected the value 4 and only one scored the value 3. The average score is 4.1 with a variation of 0.47, which means that there is consistency in the responses of the trainees who participated in this survey, and they consider that robotics has contributed to improving the students’ ability to learn (see Fig. 11).

Fig. 11
Fig. 11
Full size image

Contribution of robotics to student learning.

Finally, when trainees were asked if robotics should be implemented at school, we had interesting answers, such as “it improves logic noticeably”, “it opens doors for students”, “it helps in other subjects”, “it contributes to the creation of logical thinking” and “the children’s ability to iterate gains strength in the classroom”. We have compiled the answers and presented them in Table 6. For this alternative, we can see that the opinion of the trainees corroborates with the thought of the teachers, when they perceive the benefits of implementing robotics for student learning.

Table 6 Trainees responses on the implementation of robotics at school.
Full size table

Through the application and analysis of the questionnaires, it was possible to see that robotics influenced, motivated, and contributed to the development of the teaching-learning process of students with difficulties or disorders, according to the analysis of the results shown in Figs.  2 (students), 6 (teachers) and 11 (trainees) and the notes in Tables 4 and 6, which illustrate these points. We note that these data refer to students which have been identified as having learning difficulties, which we used to verify our hypothesis.

We can see that the results obtained so far corroborate our hypothesis that educational robotics can indeed be used, either as the main learning technique or even as an adjunct, in the process of improving the learning of students with any learning difficulties. Whether these difficulties are caused by a disorder with a well-defined ICD code, or even by any difficulties that students may encounter in their day-to-day lives at school.

Therefore, the experiments carried out and the data collected point to evidence that educational robotics can help in this process. Either as the main tool (teaching robotics) or as an adjunct (auxiliary tool in the process of teaching other subjects). However, we would also reiterate that the various challenges pointed out by teachers and trainees, whether it be the lack of structure, accessible materials, lack of time for lesson planning or even basic student difficulties, need to be considered and better investigated if we are to achieve relevant and promising results in the teaching and learning process.

Conclusion

In this paper, we conducted a mixed methods investigation of the role of educational robotics in supporting the learning of students who experience difficulties or disorders that affect their academic performance. These challenges may arise from simpler social issues or from medically recognized conditions, such as Autism Spectrum Disorder (ASD), Attention Deficit Hyperactivity Disorder (ADHD), and dyslexia, among others. The quali-quanti analysis of questionnaires completed by students, teachers, and trainees, as discussed in “Results and discussion”, confirmed our initial hypothesis, providing some new evidence that educational robotics can contribute to improving the learning performance of students facing such difficulties or disorders.

Our initial literature search in the form of a systematic review showed that there are few studies that focus on the use of robotics as a tool to help students with learning difficulties20. This was later corroborated by a broader literature search, carried out in the course of the work and presented in “Related works” Most part of the previous works are related to specific learning and do not deal with general learning disorders or difficulties, as is the case here. Initially, we found only 28 scientific articles on our subject. However, we found that only two articles specifically dealt with learning and learning difficulties. Subsequently, we identified 27 theses and dissertations. Similarly, only two of the theses directly dealt with learning and learning difficulties (one of them related to one of the articles found).

So, based on the evidence that is substantiated by the results presented here, we can now suggest that the pedagogical use of robotics can be taken as a direction to follow to improve the learning of basic skills and even the learning of transversal contents, which are sometimes a source of difficulty even for children without any learning difficulties. Furthermore, we present here a methodology focused on an interesting qualitative-quantitative approach, based on surveys, which could be used in this work to find such evidence in research in the field of educational robotics.

Thus, we suggest that the pedagogical use of robotics can serve as a promising pathway to enhance learning of both basic skills and areas of transversal content, which can pose challenges even for students without learning difficulties. In addition, the methodology based on the qualitative-quantitative approach and relying on surveys proved effective in gathering such evidence and may serve as a valuable reference for future research.

Hence, with this work, we have moved a step forward towards showing evidence that educational robotics can play a role to promote gains for teachers and trainees and for students with learning difficulties or disorders, in the general case. The results obtained so far are important in this regard. Future studies will continue to be carried out to ensure that these initiatives can be improved during their implementation. As future work, some more fine-grained checks are needed, with a view to reaching a consensus among the teachers who believe that educational robotics did not play such an important role (minority) and identifying what the students found easiest, as this can be used to the advantage of the teachers and trainees. In fact, we plan to include a control group (other students who do not present explicit difficulties) that has already participated in robotics classes. This was not included in this work because we believed that it was beyond the scope of this article. We intend to conduct further research, including statistical measurements of more variables, including observed advances, with learning rates achieved by both groups. Here, we just provided this initial observed evidence of the benefits of educational robotics for learning difficulties using a qualitative and quantitative approach.