The Role of Internet of Things and Security Aspects in STEM Education

Abstract

In the last decade, Science, Technology, Engineering, and Mathematics (STEM) education has rapidly evolved and increasingly makes use of innovative technologies. This paper aims to explore and analyze the research on integrating the Internet of Things (IoT) within STEM education and outline key aspects and emerging trends. A complete picture of the recent ten years is gained by gathering bibliometric data from the Scopus and Web of Science scientific databases. Two search queries combining IoT, STEM education, and security were submitted to extract relevant publications and obtain insight into the explored area. The manual content analysis of the study results and publications outlines several key roles of IoT in implementing STEM educational practices, which are conceptualized to reflect user viewpoints. Widespread IoT applications in STEM at different educational levels are discussed and summarized. The special focus on security aspects showed that they are underrepresented, evidenced by the small number of publications related to IoT in STEM education. The importance of including topics aimed at designing and implementing secure IoT applications in STEM-oriented curricula and courses is also discussed.

1. Introduction

Over the last decade, digitalization and various information and communication technologies (ICT) have become an essential part of the educational context to facilitate the teaching process and provide enhanced learning interactions. Much research has already outlined the necessity for adoption of innovative technology-enhanced teaching approaches to align with the digital generation [1,2]. Also, in recent years, the Internet of Things (IoT) appears to be the next-generation communication infrastructure, which allows plenty of sensors and devices to be integrated seamlessly for various implementations [3,4] and has rapidly penetrated all socio-economic areas, including industry, computing, and communication, and education does not fall behind. This technology finds various educational applications, particularly in Science, Technology, Engineering, and Mathematics (STEM) education, where it can be utilized to full extent.

STEM education can be considered an umbrella term that embraces different fields across these disciplines [5]. Most researchers consider STEM education to be an integrated learning approach comprising several STEM disciplines instead of focusing on individual ones [6,7]. STEM education is intended to prepare students with 21st-century skills relevant to technological and scientific innovations. The skills necessary for the 21st century are essential in the contemporary digitally transformed work environment. Some of these skills (e.g., digital literacy, computational thinking, critical thinking, design thinking, problem-solving skills, creativity, and collaboration) are considered to be acquired in STEM education by applying multi- and interdisciplinary approaches to teaching and learning [8]. Hence, students can develop these skills and their practical applications in solving real-world problems in STEM-oriented tasks using interdisciplinary and transdisciplinary knowledge and approaches [9]. Various IoT devices can support these approaches enabling the construction of scientific knowledge in different disciplines.

IoT appears to be a promising tool to enhance and transform the teaching-learning process in STEM [10,11]. A general introduction to IoT with a focus on related application areas, technologies, and architectures may include aspects of sensor networks, data processing and management, software, cost-effectiveness, interoperability, security, and privacy [12]. These enabling aspects serve as important decision criteria or implementation requirements for all application areas, including STEM education. The IoT architecture, from a functional point of view, may be divided into several layers—a sensor layer, a network/communication layer, a service support layer, and an application layer. Technical, semantic, and organizational interoperability must be guaranteed. A thorough systematic literature review [13] explores the benefits and challenges concerning the integration of IoT into educational settings and summarizes the implemented tools. It also reveals several research gaps in the area. Several aspects of the integration of IoT in the educational context are briefly presented in [6,14], giving a concise introduction into the complexity and interdisciplinarity of the IoT and its versatility with regard to potential application fields. These characteristics make IoT suitable for use both as a potential building block of intelligent educational systems and as a multidisciplinary subject for study in its own right.

Security is an important aspect of IoT and it has aroused research interest in recent studies [15,16]. Its characteristics present specific challenges that need appropriate solutions. IoT networks with a multitude of heterogeneous devices, having various and limited resources, produce great amounts of data that need to be protected. There are modern security approaches to the Internet of Things, such as federated learning. Many publications are devoted to IoT security in general, e.g., [17], but the focus of this study is on IoT applications in STEM. The latter topic appears to be underrepresented in the literature, and one of the research objectives of this paper is to explore the interest of the STEM-involved community in this subject.

Many researchers advocate the view that there should not be debates on whether technologies can be used, but on how they should be used to support high-quality teaching. For teachers, it is vital to know the roles of technologies employed in education, which is especially true for teaching STEM subjects. Authorities also have to be aware of the rising trend in technology integration in educational settings to provide adequate policies. Various technologies applied in the teaching-learning process can engage and empower students, thus stimulating them to be active, creative, and knowledgeable. Therefore, it is necessary to investigate the issues related to the application of IoT, one of the most ubiquitous technologies, in STEM education, benefiting educators, curricula developers, and policymakers.

The current study aims to explore and analyze the research on integrating IoT into STEM education and outline key issues and emerging trends. It is achieved by a bibliometric review based on gathering bibliographic data about publications in two of the most renowned scientific databases—Scopus and Web of Science (WoS). This bibliometric study follows the approach of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [18]. By using the combined query that concerns the application of the IoT into STEM education, relevant publications over the past ten years were extracted from Scopus and WoS. The keywords in the query denote technologies in an integrated and adaptive manner.

The acquired data were thoroughly examined and processed to achieve the research goal, namely to explore, critically analyze, and summarize the applications of IoT and IoT security aspects in STEM education reflected in scientific publications. This goal is related to the following research objectives: (i) providing an overview of research on the application of IoT in STEM education; (ii) presenting general statistics data on the topic; (iii) analyzing research terms and applied technologies to reveal research gaps; (iv) thoroughly analyzing selected eligible publications, and (v) identifying roles that IoT plays in STEM education, including security aspects, and tracing research prospects and trends.

Three research questions related to achieving the research goal were formulated:

RQ1: What are the overall picture, recent topics, and research trends regarding the applications of IoT in STEM education?

RQ2: What are the main roles of IoT in STEM education?

RQ3: How is IoT security included in STEM education?

The analysis and discussions related to these three research questions are presented in the following sections. Section 2 briefly introduces the research methods used, and the results of their application are summarized in Section 3, which presents the overall picture of the IoT in STEM education. Section 4 explores and summarizes the role of IoT in STEM education through a qualitative analysis of selected publications. The findings are conceptualized in Section 4.2 and Section 4.3. Security aspects are discussed in Section 4.4. A discussion of the findings and analysis of the answers to the research questions is presented in Section 5. The paper concludes with Section 6, which summarizes the main contributions and outlines some directions for future work.

2. Materials and Methods

The current survey aims to fill an essential gap in the existing research by combining pedagogical, technological, and security viewpoints into an integrated examination of the role of IoT in STEM education. In pursuing this goal, the research applies a mixed approach methodology when conducting the study. First, two queries regarding the role of the Internet of Things in STEM education and related security aspects were submitted to two of the most comprehensive and reputable scientific resource databases—Web of Science and Scopus. Both reflect peer-reviewed publications, have a relatively broad interdisciplinary scope, and index various types of research, including conference proceedings, journal articles, and books. These factors outline the primary reason the authors conducted the bibliometric query by extracting relevant information from these sources. The bibliometric study gives valuable insights into the topic of interest. It is not a thorough literature review but rather a snapshot of the research interest in the explored area. It enables obtaining statistical data of trends about publications, citations, and other indicators. The returned result was processed using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology to ensure comprehensive reporting of a systematic quantitative review of recent research on the explored topic [19]. Systematic literature reviews are generally considered to give a reliable approach to analyzing findings in technology-supported educational research [18]. This study uses the following procedure in the conducted literature reviews: document identification, screening, eligibility assessment, and data analysis and synthesis. The selected papers are processed using bibliometric analysis, which is done in the RStudio 2025.05.0 environment and through the application bibliometrix:biblioshiny [20]. Last, the authors reviewed the relevant publications with the highest citation rating to gain complete insights into the topics concerning the role of the Internet of Things in STEM education and related security aspects.

The obtained results suggest actionable insights for school and university teachers, policymakers, and technology professionals in providing smooth and secure IoT integration in the STEM educational practice. By considering not only pedagogical aspects but also a broader context, this research seeks to present a holistic view of the role of IoT in STEM education.

3. Overall Picture Regarding the Application of IoT in STEM Education

The scientific interest in the application of IoT in STEM education is investigated by searching two scientific databases—Scopus and Web of Science (WoS). They are believed to contain rich information about research publications worldwide and index papers from various scientific areas, especially in computer science and engineering. This paper explores the joint application of engineering and educational technologies and needs information about their interaction in the field of education. Therefore, Scopus and WoS scientific databases are good sources to obtain the desired data.

Two bibliometric queries were conducted in these databases on 17 March 2025, to explore scientific publications for the last ten years. Additional filters were “only publications in English” and “without retracted papers and editorials.” The information is retrieved from titles, authors’ keywords, and abstracts of the documents. The dataset obtained from the first search query (SQ1) is subjected to further processing using PRISMA methodology to extract unnecessary information. This process is described in detail in Section 3.1. The second search query (SQ2) is applied to the resulting dataset from SQ1 and PRISMA methodology.

The role of IoT in STEM education and the security aspects of this interrelation are explored by the following search queries:

SQ1 (IoT OR “Internet of Things”) AND (STEM OR “Science, Technology, Engineering, and Mathematics”) AND (education * OR teach * OR learn *)

SQ2 (IoT OR “Internet of Things”) AND (STEM OR “Science, Technology, Engineering, and Mathematics”) AND (education * OR teach * OR learn *) AND secur *

The first search query concerns the presence of IoT in STEM education; hence, “education *”, “teach *”, and “learn *” are included in the search. The second search query aims to outline publications that specifically explore security in the studied context. It filters the dataset of SQ1 by adding “secur *” in the query.

3.1. Acquiring the Studied Dataset Using the PRISMA Approach

The PRISMA approach is applied to the dataset for SQ1. The data for the figures below is obtained by combining the publications exported from Scopus and WoS databases, under the SQ1—346 documents from Scopus and 250 papers from WoS (Figure 1).

The number of duplicated documents was 168, and they were extracted from the combined dataset, resulting in a selection of 418 documents. These were screened for relevance based on their titles, authors’ keywords, and abstracts. The criteria used to determine which publications to include and exclude from the dataset and analyze further are presented in

Table 1. Inclusion and exclusion criteria for the relevance of the publications to qualitative analysis.

Type of Criteria	Description of the Criteria
Inclusion criteria	Topics, such as Internet of Things (IoT); Science, Technology, Engineering, and Mathematics (STEM); teaching-learning process in STEM
	IoT use in STEM educational approaches and instructional strategies
	Using IoT to make students aware of contemporary issues related to sustainability, energy efficiency, ecology, society, etc.
	IoT applications as a learning tool in STEM
	Security aspects of IoT in STEM
	Design of STEM courses using IoT
Exclusion criteria	Containing the verb “stem” in the title, abstract, or authors’ keywords
	IoT applications not in the educational context
	Aspects not focused on using IoT in STEM education (e.g., gender issues, business, and economics)
	General security aspects not related to IoT in STEM
	Citation index less than 9

.

Manuscripts that only mention the topics of interest without studying them were excluded. The resulting dataset of 215 papers was subjected to abstract analysis. To avoid irrelevant publications that mimic STEM-related topics by using the verb “stem” in the title, abstract, or keywords, these were excluded from the studied dataset. Some of the manuscripts dealing with IoT-security topics discussed issues outside of the educational context. They were excluded after a full-text analysis.

The exclusion criterion, “less than 9 citations per paper,” was chosen based on the citation data for the SQ1 dataset. The overall number of citations is 839 for 141 publications. The average number of citations per paper is approximately 6. Documents with more than 6 citations represent 29.8% (42 papers) of the cited documents (19.5% of all papers). The citation index of 6 was deemed insufficient. Therefore, a more relevant cut-off threshold was considered. For a paper to be regarded as influential, a citation index ≥ 9 per paper is needed (a 50% higher value than the average number of citations per paper). Documents with this citation index represent 11.6% of all publications and 17.7% of the cited publications. As a result, the most influential 22 publications were selected for qualitative analysis.

3.2. Bibliometric Results from SQ1

Results obtained by applying the PRISMA methodology to the combined datasets from Scopus and WoS were further processed with the software tool bibliometrix:biblioshiny [14] to analyze the data in publications’ titles, authors’ keywords, and abstracts. The bibliometric study aims to achieve deep insight into the explored topic, using quantitative techniques that help identify current research areas and trends, knowledge gaps, and future research directions.

Recalling SQ1: (IoT OR “Internet of Things”) AND (STEM OR “Science, Technology, Engineering, and Mathematics”) AND (education * OR teach * OR learn *)

Main information about the datasets for SQ1 and SQ2 is presented in

Table 2. Main information about the datasets for SQ1 and SQ2.

Description	Results for SQ1	Results for SQ2
MAIN INFORMATION ABOUT DATA
Timespan	2015:2025	2016:2025
Sources (Journals, Books, etc.)	164	21
Documents	215	24
Annual Growth Rate %	−2.84	4.61
Document Average Age	4.23	3.67
Average citations per document	3.902	14.04
DOCUMENT CONTENTS
Keywords Plus (ID)	806	118
Author’s Keywords (DE)	610	116
AUTHORS
Authors	643	91
Authors of single-authored docs	24	2
AUTHORS COLLABORATION
Single-authored docs	28	2
Co-Authors per Document	3.7	3.79
International co-authorships %	10.23	12.5
DOCUMENT TYPES
article	48	9
article; early access	1
book	2
book chapter	9
conference paper	82	6
proceedings paper	71	6
review	2	3

.

For the searched period (2015–2025), 215 publications corresponding to SQ1 are obtained, most of which are conference and proceedings papers and articles authored mainly in collaboration. International co-authorships are low, only 10%. A possible reason may be each country’s specific policies regarding educational and collaboration objectives.

Judging by publication average age (4 years, Table 2), more recent publications prevail. The plot in Figure 2, which shows the annual number of scientific publications during the period, confirms this observation. The number of publications is not an additive value but the exact amount for the respective year.

There is a clear trend of an increase in the number of publications, and it can be expected 2025 also to show a comparable amount of publications. This demonstrates a growing interest in IoT application in STEM among the research community.

The popularity of publications for SQ1, however, is relatively low, almost 4 average citations per document (Table 2). Distribution of the average citations’ number per year for 2015–2025 is shown in Figure 3.

The numbers from Figure 3 can be more clearly explained by the data in

Table 3. Annual total citation regarding all publications in the SQ1 dataset for 2015–2024.

Year	MeanTCperArt	N	MeanTCperYear	CitableYears
2015	2.50	4	0.23	11
2016	7.10	10	0.71	10
2017	7.31	13	0.81	9
2018	5.88	16	0.74	8
2019	4.48	27	0.64	7
2020	5.85	27	0.98	6
2021	4.60	25	0.92	5
2022	4.30	27	1.07	4
2023	1.22	23	0.41	3
2024	0.78	40	0.39	2

, where the mean total citation per article (MeanTCperArt) is given along with some additional data for the articles (annual article count, N, average citations per year, MeanTCperYear, and citable years). Figure 3 depicts the mean total citation per year (column four of Table 3). The papers from 2020–2022 seem to have the most impact, considering the mean value of citations per year.

The five most active countries publishing on the topic of IoT and STEM for 2015–2025 are the USA, Greece, China, Spain, and Italy (Figure 4). This ranking is made considering the authors’ affiliations, as provided in the source databases Web of Science and Scopus. Among European countries, Greek researchers are most active in launching projects concerning the IoT applications in the educational area. Although the numbers for each year’s production are additive, i.e., obtained by adding the current year’s value to the accumulated production, it is clear that there is a steady interest in the application of IoT in STEM education. The leading countries have increased their scientific production since this topic has gained more popularity in recent years.

The landscape of IoT and STEM-related publications is broadened when looking into the list of most cited countries (Figure 5). In addition to the countries with the biggest number of papers (such as the USA and Greece), there are other countries among the most cited ones—Ireland, Australia, Estonia, United Kingdom, Malaysia, and Switzerland, spanning four continents.

The dataset for SQ1 contains 610 authors’ keywords, as determined by bibliometrix:biblioshiny (Table 2). The most relevant of them are shown in Figure 6. Since SQ1 addresses IoT and STEM in the publication titles, authors’ keywords, and abstracts, it is no surprise that these words are present in different forms (e.g., Internet of Things, IoT, and Internet of Things (IoT)). Figure 6 exactly reports the authors’ choice of keywords, as listed in their publications. More in-depth analysis of the SQ1 dataset was conducted to obtain more informative outcomes for the prevailing terms.

Drawing insight from the author’s keywords (Figure 6), the dataset for SQ1 is searched for additional keywords associated with STEM education, spanning titles, abstracts, and author’s keywords. The results are summarized in

Table 4. Most frequent words in titles, abstracts, and author’s keywords in the SQ1 dataset for 2015–2025.

Text in Title, Abstract, and Keywords	Count	Documents with the Terms	Percentage of Documents
engineering	364	129	60.00%
process	113	66	30.70%
sensor	114	52	24.19%
robot	207	46	21.40%
hands-on	85	43	20.00%
programming	110	37	17.21%
Arduino	98	32	14.88%
artificial intelligence	48	29	13.49%
microcontroller	28	16	7.44%
temperature	21	15	6.98%
project-based	25	15	6.98%
raspberry	23	14	6.51%
ubiquitous computing	19	14	6.51%
hands-on learning	21	13	6.05%
project-based learning	21	13	6.05%
machine learning	23	12	5.58%
computational thinking	23	11	5.12%
kits	20	10	4.65%
educational robotics	18	10	4.65%
virtual reality	16	9	4.19%
remote laborator* (-y, -ies, etc.)	14	8	3.72%
augmented reality	15	8	3.72%
experiential learning	15	7	3.26%
humidity	10	5	2.33%
smart schoolhouse	19	5	2.33%
online learning	7	4	1.86%
automatic control	3	2	0.93%

, where the different forms of the terms are merged, and terms are ordered by percentage of the documents that contain them. As shown in the table, STEM-related terms occupy the leading positions. Terms with a presence of more than 10% in the publications are associated with engineering (process, sensor, robot, programming, computer science, Arduino), intelligent technologies (artificial intelligence), and engineering education (hands-on, engineering, laborator*). That is expected since the search focus is on STEM education and IoT applications, which are both technical subjects.

Trend topics in publications from the SQ1 dataset for 2015–2025, based on authors’ keywords, include artificial intelligence, project-based learning, hands-on learning (for the last three years), Arduino, sensors, programming, robotics, etc. (Figure 7). The greater the term’s frequency, the bigger the respective mark. Terms related to our search are obviously among the trending topics, but it is more substantial to see the context in which they are used because this can provide some insight into the research directions. A more in-depth analysis of selected publications is presented in Section 4.

3.3. Bibliometric Results from SQ2

One of the objectives of this study was to explore the security issues related to applying IoT in the STEM context (RQ3). Therefore, the dataset extracted by SQ1 was searched for security-related publications. The results obtained after submitting SQ2 to the larger SQ1 dataset contain 24 documents (Table 2). Only 11% of the papers in the SQ1 dataset (24 out of 215 publications) discuss security issues in the studied area. This outcome may be due to minor interest in the subject or may imply a promising field for future research. Security-related publications on IoT in STEM education encompass a wide range of types, including articles, conference and proceeding papers, and reviews. Their average citation count per document is 14 compared to 4 for SQ1 documents. Publications with multiple authors are prevailing.

Although the dataset resulting from SQ2 is comparatively small (24 documents), there can still be seen a growing tendency in the annual number of publications (Figure 8). The topics related to IoT security in STEM context are underrepresented within the scientific community yet. A deeper investigation is needed to reveal research gaps and prospects.

The first six countries with the biggest number of publications concerning IoT security in STEM education are the USA, Spain, China, United Arab Emirates, Australia, and India (Figure 9), demonstrating a broad geographical span. The number of publications is aggregated for the period.

The most relevant words in the dataset for SQ2, according to bibliometrix:biblioshiny, are listed in Figure 10. Again, they are based on the authors’ keywords. Apart from the obvious terms Internet of Things and STEM education, there are terms concerning security approaches such as federated learning, intrusion detection, and data models.

Trend topics for SQ2 are federated learning, intrusion detection, and cloud computing. Federated learning and intrusion detection are among the recent trends, as they concern the security of IoT networks. Cloud computing and hands-on learning are also mentioned in the STEM education-related topics. In addition, terms such as blockchain, game theory, hardware, and servers are among the authors’ keywords in the publications for SQ2. Their occurrences, however, are comparatively rare and not sufficient to draw conclusions.

In summary, bibliometric data derived from the dataset for SQ2 provide some insight into the research interest in security aspects of IoT in STEM education; however more profound examination of the publications is needed. The examination of the publications in the dataset for SQ2 outlines security approaches used in IoT networks and their application in STEM disciplines. A more in-depth analysis of the SQ2 papers is presented in Section 4.4.

4. The Role of IoT in STEM Education

Based on the bibliometric survey, the most influential publications on the researched topic are identified. To gain a complete understanding of the role of IoT in STEM education, these publications are further critically analyzed. A comprehensive view and outline of the findings are presented in the following subsections.

4.1. Analysis of the Application Contexts

In the following paragraphs, a qualitative analysis of some influential papers from SQ1 dataset is carried out to obtain relevant answers to the research questions outlined at the end of the introduction. The criterion for paper influence chosen is that the papers should have a citation count of 9 or more.

The research presented in [21,22] attempts to raise students’ awareness about energy efficiency and sustainability through lab exercises. The combination of IoT and augmented reality (AR) is used as a setup in [21] to provide students with easy access to sensor data about energy consumption. The authors use their own kits based on Arduino and Raspberry Pi to teach students how to gather environmental parameters and measure energy consumption. The AR implementation uses overlays, three-dimensional building models, and QR codes to augment a camera video stream (shown on a tablet) with digital data. The data includes information about IoT devices and available sensors on them, real-time values from the sensors, and debugging information.

Another way to achieve the goal of increasing students’ motivation and engagement in energy saving and sustainability is to apply a combination of gamification and IoT [22]. In the cited project, IoT devices are used in data-driven competitive group-based educational activities to monitor the environmental parameters in a classroom or a building and then make students aware of the impact of their personal behavior on the energy consumption. Cloud infrastructure is used to process real-time data and make it accessible to mobile clients via companion applications. The gamification concept is implemented via an online application and offers multiple online quests that can be supplemented through real-world class activities. Sharing experiences and achievements via community areas and blogs is encouraged. A peer-based grading system is introduced to stimulate active participation. Emphasis is placed on computer science, but other STEM disciplines are also mentioned. The study highlights the role of competitiveness, the benefits of creating short and interesting materials (including IoT kits), and the importance of close collaboration with educators and providing support to them.

In summary, IoT proves to be modern and interesting. It may increase the motivation and participation level of both educators and students. It also provides the basis for raising awareness and making informed decisions about topics of public interest, e.g., energy consumption.

In a study concerning ecological issues, Tziortzioti et al. designed a water sensor IoT kit based on Arduino to measure different parameters of seawater (temperature, oxygen, pH, etc.) and involve students in active learning [23]. The students can participate in their own projects, collecting and analyzing data, and making conclusions regarding coastal water quality. In this way, they obtain knowledge in the areas of science and technology and use it to solve real problems.

Different hands-on education activities are discussed in [24,25]. A hands-on approach to teaching IoT, cloud computing, and blockchain technologies is presented in [24]. The main goal is to close the perceived gap between traditional courses and emerging industrial technologies by introducing hands-on laboratory exercises. This approach contrasts with traditional lecture formats that are perceived as boring and incapable of motivating students, whose dropout rates from STEM majors are already high. The need for continuous updates of the course materials and the different student backgrounds pose challenges. Thus, educators may benefit from sharing their experiences and adopting best practices. Other challenges are related to the scarcity of integrated study materials, the lack of hands-on laboratory materials (especially on massive open online course platforms), the interdisciplinary nature of embedded systems, and the short semester time (16 weeks). Besides the Raspberry Pi hardware board and a camera sensor, cloud-based software (Jupyter Notebook) is used for Python programming. The course is supplemented by cryptographic concepts and blockchain-related topics. Student feedback is positive. This study outlines a trend for raising the motivation and engagement of students by organizing hands-on laboratory courses about IoT and blockchains. The main role of IoT is to provide an illustration and application of electrical engineering and computer science knowledge in an interesting and modern setting. Security and cryptography are also touched on. One potential danger is that this topic differs from mainstream uses of embedded systems and IoT, so the course may not be representative of widespread IoT use cases.

In [25], students’ motivation towards STEM education is raised through IoT hands-on education activities. The goal is to make students acquainted with career choices and job opportunities in the STEM field. IoT is used both as a learning aid and a learning objective. Systematic preparation of materials and technical equipment, planning, documentation, support, and skill enhancement of teachers, as well as student group formation, are critical for success. Educational materials include detailed descriptions, worksheets, codes, presentations, images and videos, wikis, forums, and blogs, as well as an Arduino-based hardware experimentation kit, including a central board and a set of sensors and actuators. The goal of the practical exercise is to implement a smart recycling bin. The student feedback is mostly positive. According to the authors, short-term educational scenarios are preferred to long-term ones, and boys are more satisfied than girls. Teachers’ feedback is also positive, emphasizing the acquisition of new skills and the increase in their confidence in using technological equipment. Problems are caused by the low teacher-to-student ratio, the different backgrounds of students, and the limited time available for equipment preparation and participation in the educational scenario. After the interventions, the willingness to choose a STEM career path increases for girls and decreases for boys. This outcome is presumably due to acquiring a better understanding of STEM career paths and the perceived capability to follow one. In conclusion, the study uses IoT as a motivating element for students to make them better acquainted with future STEM career paths. A particular focus is placed on the reactions and attitude changes of girls vs. boys, emphasizing gender differences.

The authors of [26] suggest developing a module course on Maker Education. The goal is that 5th-grade learners form a preliminary cognitive concept of the Internet of Things. The study collected and analyzed learning data and assessed the learning effectiveness of this course, while considering the factors of “Theoretical knowledge,” “Practical application,” and “Maker’s mindset.” The research results show that the students have acquired and internalized the STEM theoretical knowledge during the courses, while transforming it into practical actions. Thus, it appears that Maker education applied in STEM subjects allows students to learn how to integrate and apply knowledge through hands-on activities and projects.

Kamal et al. [27] present a learning kit that is specifically created to facilitate students’ understanding of IoT and blockchain and to ignite their interest in learning these technologies. The proposed approach considers the context of project-based learning. The learning kit is realized in three parts: data collection from sensors and Arduino Uno, data transfer to Raspberry Pi, and usage of a cloud service with a suitable interface to present results. The students are capable of understanding how communication between Arduino, Raspberry Pi, and cloud platforms happens, as well as how to design a cloud platform.

Benita et al. [28] discuss the developed smart learning ecosystem to stimulate data-driven thinking of children and adolescent students in the context of experimental and collaborative learning. The smart learning ecosystem comprises various IoT devices (sensors for data gathering), cloud computing (data processing), and analytical services (data visualization). It gives learners the possibility to make experiments with data and to obtain its meaning.

Cornetta et al. [29] introduce the idea of fabrication-as-a-service (FaaS) based on IoT and Industry 4.0 concepts to give an opportunity for remote access to schools and universities and implement the concepts of learning by doing. This approach leads to improved usage of Fab labs for educational purposes and focuses on knowledge sharing and cooperation between students and teachers. The architecture of FaaS includes a cloud-based hub and Fab labs connected in a network. Two use cases are demonstrated: the usage of FaaS for the collaborative working of several groups with students and teachers on one project, and FaaS usage by two labs to share an assembly line in electronics.

Glaroudis et al. [30] create a toolset that integrates equipment for IoT, mobile and ubiquitous computing, the educational environment for collaborative interaction, and suitable scenarios for the realization of active, problem-driven, and collaborative learning. The proposed approach introduces students from secondary school to IoT and enables their understanding of several issues in science, engineering, math, and technology. The purpose is for the students to obtain knowledge regarding IoT, mobile, and ubiquitous technologies and to strive for a career in this area.

Kusmin et al. [31] present a view of a smart schoolhouse that is based on sensor kits (environmental sensors, cloth and body sensors, digital art sensors) suitable for experimentation by school students. Thus, important and real problems could be solved through applying methods like inquiry and project-based learning. Such an approach facilitates the integration of IoT concepts in teaching scenarios and teaching materials and stimulates students’ interest in technology and engineering.

A literature review presented in [10] focuses on the curriculum, pedagogy, and assessment aspects related to IoT-based educational content. It uses a five-element model containing a sensing layer, a network layer, a service layer, an interface layer, and an overarching cybersecurity block. The authors report that one key focus is on K-12 students who learn about the functioning of the sensing layer via low-cost sensor devices and electronic kits. Key activities include interfacing with digital and analog environmental and proximity sensors, RFID integration, working with biomedical sensors, and gas monitoring. Single-board computers such as Arduino and Raspberry Pi are widely employed. Educational content for the network layer overlaps with computer networking courses. It includes Wi-Fi & Ethernet, Bluetooth, Zigbee, and LoRa. Cloud services also play a significant role in the service layer curriculum. The curriculum for the interface layer deals with user interface design concepts and testing for verification and validation. Light-emitting diodes (LEDs), liquid crystal displays (LCDs), web-based dashboards, and mobile and voice interfaces are used. Programming languages C/C++, JavaScript, Lua, and Python are employed. The learning approaches that are often used include problem-based learning, project-based learning, activity-based learning, inquiry-based learning, and collaborative learning. Remote laboratories are also a popular topic. The assessment includes video presentations, code files, reports, quizzes, and essays (formative assessment), as well as presenting results about final projects (summative assessment). Student feedback is also collected, and interviews are conducted. Real-world projects are presented that include home automation, weather and air quality monitoring, and wearable devices. The broad review reveals an emphasis on low cost and a surprising lack of industrial, automotive, or mission-critical educational content and hands-on projects. A tendency is shown to regard IoT as a supplementary topic and a part of general computer science, thus losing part of its multidisciplinary nature. One additional insight of the study is that cybersecurity has only recently started gaining popularity.

The paper [32] presents a meta-analysis of the effects of Arduino- and Scratch-enabled interventions to provide aggregated evidence of the impact of open-source Arduino and Scratch on students in STEM disciplines. The results show that the low-cost, simple, easy-to-use solutions have an overall positive effect on students’ STEM academic achievement and introduce them to computational thinking-focused engineering activities.

Byrne et al. [33] report on a constructivist 21st-century learning model to implement a week-long workshop (a hackathon) to encourage pre-university students to pursue careers in STEM and computer science, in particular. Participating students were involved in a process of setting up, investigating, planning, creating, and presenting a prototype of a wearable/IoT device while working in groups. Summarizing their experience, they answered two questionnaires—before and after the participation in the workshop. The findings indicate that the applied 21st-century learning model combined with a hackathon activity can effectively motivate and increase the self-efficacy of pre-university students in subjects related to wearables and IoT.

A similar study [34] explores the use of tangible toolkits for physical computing to support pedagogies of collaboration and production. The focus is on learning through IoT about STEM, specifically computer science and engineering. The study uses mixed methods—video and audio recordings, interviews, observations, surveys, collaborative designs, a hack event for collaborative and problem-based learning, etc. The outcomes confirm the importance of IoT as a technology-enhanced learning tool supported by its flexibility and ability to put the learners in direct contact with algorithms and sensors so that theoretical knowledge becomes concrete and realizable.

Another motivational and career-seeking role of the Internet of Things is studied in [35]. A one-day female-focused STEM-based entrepreneurship program is conducted, and data is collected and analyzed from two surveys to explore the entrepreneurial intentions of secondary school female students. The program is aimed at encouraging teenage girls to develop and implement computational solutions using IoT. The findings show that entrepreneurial attitudes in young female students are associated with soft-skills acquisition, particularly in problem-solving, creative thinking, risk-taking, and leadership development. Positive role modeling and peer-to-peer learning were also considered key factors in encouraging entrepreneurial intent.

Porter et al. [36] present project activities focused on stimulating interest in STEM and promoting STEM careers by encouraging students to gain knowledge and skills in engineering. It is achieved by interacting with practicing professionals on authentic experiential learning activities and engineering design projects related to building automation. A basic building automation platform applicable to secondary education is offered. It includes a programmable IoT Building Monitoring Device based on the Texas Instruments LaunchPad, which is used as a simple building block for the STEM educational modules. Programs for the various learning modules can be provided, where teachers can edit and modify them to implement new functions and capabilities. Thus, a building monitoring device can be developed to observe and collect data about several environmental parameters and transmit them to the cloud for additional processing and analysis.

Authors of the research [37] present an approach for system reconfiguration to allow different scenario configurations depending on the educational needs. The system is built of modules, peripherals, and external elements that can be organized according to the educational scenario. The system architecture is designed of different modules (based on Arduino Uno), which can be easily interchanged. An FPGA is central to the design and allows hardware detection and reconfiguration. The approach and the architecture support the inclusion of robotics in a collaborative development environment to promote the innovation and motivation of the students during the learning process within a STEM context.

The study presented in [38] explores the teaching and learning of IoT-enabled STEM subjects in online classrooms during the COVID-19 pandemic. It reveals teachers’ experiences in this context, their selection of particular Web 2.0 tools to establish online classrooms, and their pedagogical practice in adapting the hands-on activities to suit online teaching, especially concerning experiments in science laboratories. Hence, technology-enhanced learning allows students to be engaged via online tutorials and virtual learning environments in remote experimentation strategies in STEM activities related to the real world. The IoT allows for the integration of various technological and communication tools that enable remote education, thus requiring the technological literacy of students and teachers.

The research in [39] proposes implementing a cost-effective system that generates an airflow that can be controlled remotely from a tablet or smartphone using open-source software. The system is of realistic size and built from simple electronic components (e.g., Arduino, Raspberry Pi, IoT) and some reused electrical equipment. A project-based learning approach is applied in its creation. From an educational point of view, the proposed approach is mainly intended for agricultural engineering students, aiming to present better the basics of automatic control and several hardware, software, and network principles. Teamwork during the implementation of the proposed system allows for strengthening collaboration and organizational skills. The suggested approach aligns with the general directions for sustainability and recycling.

The study [40] collects data and examines information about courses that integrate hardware prototyping platforms into learning technologies. The findings are analyzed according to several characteristics, such as course type (i.e., STEM or non-STEM), platform used (i.e., Raspberry Pi, Arduino, BeagleBone Board, etc.), purpose (i.e., experimental or reference approach), and others. The goal is to provide an overview of the current state of integration of hardware prototyping platforms and to determine the context in which they are integrated into STEM and non-STEM education.

The authors of [41] cite IoT as one of the core components of digital transformations that can be used in smart play corners in kindergartens. It is an important topic for young children to know about as early as kindergarten. The children’s play corners are equipped with materials for pretend play, such as old laptops and wooden tablets. In this way, young children become acquainted with digital processes and foster and improve their skills for digital problem-solving. The study acknowledges IoT as a significant trend in the digital age, and familiarity with it is regarded as part of the core digital competencies needed in modern life.

The critical analysis of the investigated publications highlights the widespread IoT applications in STEM at each educational level (

Table 5. Widespread IoT applications in STEM at different educational levels.

Target Group Students	IoT Applications	Source
Kindergarten and preschool	Mockup IoT models and figures to sensitize children to the existence of various devices through pretend play	[41]
Elementary school students	Maker education, hands-on activities, and a personal project	[26]
Secondary school and pre-university students	Workshops encouraging the pursuit of careers in STEM; exploring STEM-based entrepreneurship attitudes; knowledge construction during learning by making in complex environments	[33,34,35]
	Making students aware of environmental and societal issues via school projects involving IoT	[21,22]
	Hands-on and gamification for IoT-based educational tools	[22]
	Facilitating career choice by raising students’ awareness of IoT through hands-on activities	[25,36]
University students	Hands-on laboratory-based approach using IoT, cloud computing, and blockchain applications	[24]
	Project-driven STEM education	[39]

). Most of the suggested applications are intended for high secondary school and university students. They provide setups for maker education, hands-on activities, workshops, and hack events. There are attempts to introduce IoT-oriented STEM courses at the early educational stages, thereby involving young children in more technical and practical activities. The Internet of Things plays a substantial role to motivate pre-university students to pursue engineering and science education.

4.2. Summary of Findings

The qualitative analysis of the selected most popular papers from SQ1 showed a variety of technologies used in IoT applications for STEM disciplines. The preferred technologies are usually open-source and free, but they still give good opportunities to use different pedagogical approaches. The specific STEM education context, the applied teaching-learning approach, and the technological solution used in each studied manuscript are outlined in

Table 6. Summary of the qualitative analysis of the selected papers.

Reference	STEM Education Context	Teaching-Learning Approach	Used Technological Solution
Mylonas et al. [21]	To raise the awareness of students about issues of sustainability and energy efficiency through lab exercises	Laboratory exercises	Combination of IoT, (Arduino, Raspberry Pi, lab kits) and AR visualization.
Mylonas et al. [22]	To increase students’ motivation and engagement, To raise the awareness of students about energy saving and sustainability	Learning based on competition and gamification	Combination of IoT, Cloud-based services (GAIA Platform, Apache Storm) and Internet-based applications
Abichandani et al. [10]	To train skills related to sensor data acquisition and IoT	Problem-based learning, project-based learning, activity-based learning, inquiry-based learning, collaborative learning	Low-cost IoT hardware and open-source software, mainly home-automation projects
Rao et al. [24]	To raise the motivation and engagement of students by organizing hands-on laboratory courses	Hands-on multidisciplinary laboratory exercises	IoT in combination with cloud-based services
Spyropoulou et al. [25]	To make students acquainted with career choices and job opportunities in the STEM field	Hands-on group-based education activities	IoT lab kits and educational materials
Hollenstein et al. [41]	To make young children acquainted with digital processes and develop skills for digital problem solving	Smart play corners in kindergartens	Materials for pretend play (e.g., old laptops, wooden tablets) in play corners
Kamal et al. [27]	Students to gain knowledge and skills regarding IoT and blockchain technologies	Project-based learning	Learning kit with Arduino, Raspberry Pi, and cloud platform
Benita et al. [28]	To stimulate data-driven thinking	Experimental and collaborative learning	Smart learning ecosystem- IoT devices, cloud-based services, analytical platform
Cornetta et al. [29]	To improve usage of Fab labs for educational purposes, knowledge sharing, cooperation between students and teachers	Learning by doing	Fabrication-as-a-service (FaaS)—cloud-based hub and Fab labs distributed over a network
Glaroudis et al. [30]	To give knowledge in the field of IoT, mobile, and ubiquitous computing and to strive career in this area	Active, problem-driven, and collaborative learning	A toolset that integrates equipment for IoT, mobile and ubiquitous computing, collaborative educational environment, and suitable scenarios
Kusmin et al. [31]	To improve teachers’ and students’ competences in the IoT domain	Inquiry and project-based learning	Smart school house that is based on sensor kits (environmental sensors, cloth and body sensors, digital art sensors)
Tziortzioti et al. [23]	To obtain knowledge in science and technology and use it for solving real problems	Active learning	Water sensor IoT kit based on Arduino
Fidai et al. [32]	To inform STEM instructional leaders and policy makers about the advantages of open-source Arduino hardware and freely available Scratch software in STEM education	Hands-on engineering activities for K–12 and post-secondary classrooms	Open-source Arduino hardware and free Scratch software
Byrne et al. [33]	To encourage pre-university teenagers to pursue careers in STEM, with a particular emphasis on computer science	A week-long workshop, Collaborative, technology-mediated, and project-based learning	Wearables, Raspberry Pi, Arduino Uno, robotic vehicle chassis with onboard motor controller and Grove
Merino et al. [37]	Wireless Educational Platform approach to be included in STEM educational programs for school students	Collaborative development environment to promote the innovation and motivation of the students within a STEM context	Arduino, FPGA, robots
Shahin et al. 2021 [35]	To explore the effects of a one-day female-focused STEM-based entrepreneurship program on the entrepreneurial intention of secondary school female students	Exploring entrepreneurial attitudes in secondary school female students, two surveys to assess the effect of the program	IoT component using the micro:bit device
Charlton & Avramides [34]	To examine learning by doing as a medium for students to explore STEM, with a particular focus on computer science and engineering	Learning by doing, technology-enhanced learning, collaborative and problem-based learning	IoT, sensors, Arduino, Internet
Yeh et al. et al. [26]	To apply Maker education in STEM subjects; IoT as the curriculum development topic	Hands-on activities, practical actions, project-based learnings	Webduino Smart device embedded in the Esp8266 module, Webduino Blockly software
Makamure & Tsakeni [38]	To teach IoT-enabled STEM subjects in virtual classrooms based on Web 2.0 tools	Hands-on activities suitable to online teaching, experiments in online science laboratories, Teacher-centered instructional strategies	IoT infrastructure and gadgets that enable connectivity
Porter et al. [36]	To enable interaction with practicing professionals in authentic experiential environment	Learning activities and engineering design projects related to building automation	Programmable IoT, Building Monitoring Device (IBMD), Texas Instruments LaunchPad
Loukatos et al. [39]	To present basics of automatic control and hardware, software, and network principles to students in agricultural engineering	Project-based learning approach; Teamworking	Realistic size system remotely controlled by a tablet or smartphone built by simple components (Arduino, Raspberry Pi, IoT) and reused electrical equipment
Al-Masri et al. [40]	Overview of the current state of integration of hardware prototyping platforms and the context in which they are integrated into STEM and non-STEM education	Courses that integrate hardware prototyping platforms into learning technologies	Raspberry Pi, Arduino, BeagleBone Board

.

Several STEM courses and activities use IoT as a motivational tool [25,33,34,35] for pre-university students to pursue careers in engineering and computer science. Others are focused not only on gaining knowledge in programming [32] and robotics [37] but also on acquiring awareness of more complex global issues, such as sustainability [21,22], energy efficiency [21,22], reusability [39], etc. IoT devices facilitate experiments to investigate real-world phenomena [42]. However, IoT-enabled STEM teaching and learning require an appropriate investment in infrastructure and devices to provide connectivity and functionality.

There are some quantitative estimates in the literature indicating the impact of IoT application on students’ involvement, achievement, motivation, and career orientation to STEM disciplines. The meta-analysis in [32] states an overall positive effect of Arduino- and Scratch-enabled interventions, particularly on students’ STEM academic achievement and their perceptions towards STEM. In [25], the authors report a 4–10% increase in the interest of K-12 students in technology and science after the implementation of an IoT-oriented educational scenario. A gender difference exists, for example, in the change in perspective on technology, with 20% of female students and 14.7% of male students.

4.3. Conceptualization of Findings

The summary of analyzed papers (Section 4.1) and the outline of applied technologies for the Internet of Things in STEM education (Section 4.2) demonstrated the flexibility and suitability of IoT for STEM. Its interdisciplinarity, connection to the real world, and thus to practical solutions may attract learners towards science and engineering. It can help students see how the school subjects find actual applications in meaningful settings and entice them to seek careers in the STEM field. IoT is a relevant platform for developing technical skills and acquiring digital competence while gaining awareness of ICT innovations. An initial classification of the roles of IoT in STEM education, which is derived from the critical analysis of the eligible publications, includes:

• Motivation in STEM subjects;
• Digital competence—development of 21st-century skills;
• Technical skill development;
• Relationship between theoretical science knowledge and real-world applications;
• Introduction to innovations;
• IoT as a technology-enhanced learning tool that allows iterative use;
• Multidisciplinary context;
• Awareness of societal, ecological, and economic issues;
• Career motivation and choice, as well as entrepreneurship programs.

Figure 11 presents the conceptualization of the integration of the IoT into STEM education, reflecting the main facets of the teaching-learning process, namely the actors and activities. The findings of the qualitative analysis in Section 4.1 are summarized and synthesized as blocks in the figure. The results reported in the papers emphasize the impact of IoT on the groups of teachers, students, and pedagogy based on the provided functionalities. The Internet of Things facilitates interactions between various technologies and participants in the STEM educational process. The roles of IoT in STEM education are grounded on their technological basis (i.e., variety of devices, sensors, kits, and corresponding software tools and services) and reflect the applications of IoT concerning teachers, students, and used pedagogical approaches in teaching STEM subjects.

The Internet of Things can be involved in students’ learning at the very beginning, so they have five essential roles for students: (1) Boosting motivation in STEM, (2) Enabling training skills in related areas (e.g., embedded systems and data processing); (3) Giving possibility to obtain knowledge in the related topics, such as IoT technologies, science, technologies, and engineering, etc.; (4) Gaining competence in 21st-century skills (e.g., digital problem-solving, computational thinking, data-driven thinking, collaboration) and at the end (5) Students are presented with possible job opportunities and can choose their careers.

The adoption of IoT in the teaching process benefits several aspects of teachers’ work. These aspects include stimulating professional development, enhancing competencies in IoT, gaining multidisciplinary insights, improving cooperation between students and teachers, and increasing communication with academics and industry experts in the STEM area. Further, IoT enables teachers to implement concepts from different STEM domains to design and conduct interdisciplinary and multidisciplinary lessons and courses. IoT helps present STEM knowledge not in an abstract and pure theoretical manner but related to real-world problems.

The Internet of Things can be smoothly integrated into most pedagogical approaches applied in STEM teaching. When augmented by IoT, these approaches are usually enhanced and raise their effectiveness. IoT is especially valuable for data acquisition and process control in engineering approach, experimental learning, problem-based learning, and project-based learning. Further, in hands-on activities, explorative learning, and learning by doing, students acquire specific skills in dealing with various IoT devices. In all learning approaches involving IoT data acquisition, students gain knowledge in heterogeneous data processing, data analysis, data storage, data visualization, etc. The intrinsic interdisciplinarity of IoT applications stimulates students to explore and find the relationships among STEM disciplines. Their real-world connection can motivate students to be engaged in socially important causes and pursue careers in computer science and engineering.

4.4. Security Aspects of IoT in STEM Educational Courses

The result from the second query (IoT OR “Internet of Things”) AND (STEM OR “Science, Technology, Engineering, and Mathematics”) AND (education * OR teach * OR learn *) AND (secur *) is also explored to outline the place and importance of security in implementing IoT-based STEM educational environments. The initial list of publications discussing security in the studied dataset of 215 documents represented 11% of all documents. Seventeen of them were studied and divided into two groups. The documents in the first group discuss IoT security issues, but not in the context of STEM education. The second group includes six articles representing 2.8% of the whole dataset. They indicate not only the importance of studying IoT concepts for students in STEM courses but also the importance of IoT security aspects (cybersecurity, internet security, network security).

The paper by Azad and Hashemian [43] describes several possible applications of cyber-physical systems (smart houses, robots, embedded systems) in STEM education and the related cybersecurity discipline, which considers topics related to network and web security. The first cyber-physical system is a smart house on a small scale, with the possibility of remote temperature and light control through a specially created user interface. For the realization of this project, NI LabView, an Arduino microcontroller, and sensors for temperature and light are used. The goal of the second system is to develop a mobile robot capable of being remotely controlled via a graphical interface. The solution utilizes NI LabVIEW, Arduino, an IP camera, IR sensors, and Bluetooth. The third cyber-physical system is can be programmed remotely and includes a Python platform, Arduino, an IP camera, a stepper motor, a liquid crystal display, a seven-segment display, and light-emitting diodes. The described cyber-physical systems are created by means of various hardware, software, and communication technologies, which require knowledge of different security measures to secure the components and the systems as a whole.

STEM education is intended to give learners the skills they need in the domain of industrial IoT and to prepare them for successful professional realization [44]. Learners must certify their skills in the design, development, and maintenance of secure industrial IoT systems. For this purpose, a remotely accessible hands-on lab with integrated industrial systems for automation and control is created. Such a laboratory is suitable for use in courses that are practically oriented knowledge is acquired through experimentation, combining theoretical concepts with practical exercises. The laboratory also develops a range of real-world problem-solving skills while providing opportunities for collaboration, both remotely and face-to-face. When implementing a laboratory in the field of AI, one must think about appropriate software and hardware, what communication venues to implement, as well as about potential cybersecurity risks. The authors acknowledge the importance of security in the IoT ecosystem, placing the topic of security alongside the main building blocks.

Rao and Dave [24] present a hands-on laboratory to facilitate teaching in IoT and embedded systems. The created laboratory practices address security issues in data (image) processing and storage, while the cryptographic hash algorithm SHA-256 is introduced to students. Specifically, students learn to perform comprehensive image processing, create secure records by applying cryptographic techniques such as hashing, and send and store them in a cloud infrastructure. Special emphasis is placed on a better understanding of security when working with camera applications.

Sample et al. [45] report that students’ awareness of contemporary issues and challenges in cybersecurity can be achieved by including STEM courses in the curriculum for undergraduate and graduate students. Furthermore, courses can be made available in two formats: with and without certification. Students should be taught how to design secure algorithms, code, and systems, as well as how to understand emerging cyber events in the fields of IoT and cyber-physical systems. The authors emphasize that in the field of IoT, it is especially important for students to understand the significance and consequences of cyber events. It is necessary to build a generation of defenders who have specific knowledge of cybersecurity and can also apply it in a particular environment. This generation must be able to view issues from different perspectives and think critically in order to solve challenging problems in cybersecurity.

Jamro [46] emphasizes that the key aspect of IoT security is the implementation of concepts related to quality of service and quality of experience. These concepts should be included in students’ STEM education through relevant projects in topics like coding and digital electronics. The main problems related to IoT security include the need for the development of security cultures in institutions, the integration of suitable policies in STEM courses and the lack of knowledgeable people. It is argued that security threats come from IoT manufacturers, who must produce secure IoT devices, and from end-users, who must understand how to protect the IoT solutions they use.

In a literature review, Abichandani et al. [10] discuss the importance and need for student training in IoT and cybersecurity skills. For this purpose, curriculum, pedagogies and suitable assessment approaches should be implemented through low-cost and open-source technologies. The authors come to the conclusion that providing appropriate experiences in a learning environment can help create the next generation of IoT experts. The significant role of the education system is emphasized in this regard. Some challenging issues for practical implementations of IoT are discussed. They are related to the following matters: hardware and software that should be continuously upgraded and updated, educators who should be trained in contemporary technologies and IoT landscapes, application of suitable assessment tasks to assess students’ knowledge and skills, lessons and instructions that should be constantly updated, reflecting emerging technological changes.

The explored literature shows that IoT requires secure measures in transmitting, processing, and storing collected data. Therefore, the researchers propose the development of appropriate STEM curricula, courses, and pedagogical strategies that introduce students to the issues of creating secure IoT applications. The acquisition of skills related to the creation of secure algorithms, code, and systems is of particular importance in the continuously and dynamically changing environment of cyberattacks. These findings are summarized and presented as a conceptual model to highlight the importance of studying the security aspects of IoT in STEM education (Figure 12). It demonstrates that the organization of STEM education should include dedicated courses with appropriate pedagogical scenarios and assessment strategies to encourage students’ active participation in conducting experiments and drawing conclusions. Regardless of their major, students should be allowed to participate in STEM courses that treat topics related to implementing secure IoT applications. Security issues should also be discussed in courses that end with the certification of participants, as well as in courses from the bachelor’s and master’s curricula. Building secure IoT applications requires acquiring knowledge and skills to design and develop secure algorithms, program code, and, ultimately, secure IoT systems. The issues related to the maintenance of secure IoT solutions should also be considered for exploration.

The current study reveals that the topic of security concerning IoT integration in STEM education is not sufficiently addressed and developed, however, it is extremely important, especially in contemporary cyberspace. Learners should be exposed to challenging security issues and possible solutions from an early age, even in primary school years. Teachers should strive to prepare STEM lessons that deepen students’ knowledge and skills for implementing secure approaches when using or creating IoT applications. Engineering courses should include security topics to introduce future designers of cyber-physical systems to the security challenges they would face.

5. Discussion

The goal of this paper is to explore and analyze the applications of IoT and IoT security aspects in STEM education, thus outlining the role of Internet of Things in this field. Three research questions shape the focus of our study as defined in the Introduction.

To answer RQ1, a search query (SQ1) was conducted to explore and analyze IoT applications in STEM education and IoT-enhanced learning environments for STEM education. The search query delivered a dataset of 215 publications that were analyzed by titles, authors’ keywords, and abstracts using the bibliometrix:biblioshiny application. The results demonstrate a growing interest of the research community in the application of IoT in STEM disciplines. According to the most frequently used keywords, the trending topics are in the field of computer science and engineering, such as Arduino, artificial intelligence, robotics, and programming. The IoT is used in STEM disciplines in the context of project-based learning, hands-on activities, and engineering education. The publications in the SQ1 dataset were searched further for keywords associated with STEM education and IoT. Additional topics suggest orientation towards engineering and hands-on education (e.g., sensor, microcontroller, Raspberry Pi, etc.).

In summary, several research trends regarding the applications of IoT in STEM education have been identified (RQ1):

• IoT proves to be an innovative and engaging tool, which may increase the motivation and participation level of both teachers and students in STEM education;
• The bibliometric analysis reveals that the USA and Greece are among the countries where the topics related to IoT and STEM are mostly researched.
• Popular topics are related to engineering (process, sensor, robot, programming, computer science, Arduino, robotics), intelligent technologies (artificial intelligence), and engineering education (hands-on and project-based learning).

Following the trends in applying IoT in STEM education, it was essential to investigate the specific roles the Internet of Things takes and what makes it suitable for teaching STEM disciplines. A set of 22 full-text publications was analyzed to answer RQ2. Many IoT technologies are applied in STEM disciplines and are used both as a teaching subject and a tool for teaching other STEM subjects.

Within the highlighted research trends, the role of IoT in STEM education is often manifold and includes several essential applications (RQ2):

• IoT serves as a motivational tool due to its inherent “wow”-factor—IoT devices are innovative, real, often flashy and unorthodox, which raises students’ interest, excitement, and involvement;
• IoT is used as an application example to exercise and assess competencies learned in computer science courses (e.g., programming languages, cloud services, cryptography, etc.) in a practical setting;
• IoT provides a suitable basis for working on real-world application projects. Students’ grades can reflect their performance in real problem-solving scenarios benefiting from expert knowledge;
• IoT helps improve students’ awareness of other topics and issues that have an impact on society, e.g., reducing energy consumption or managing waste through recycling. In this scenario, IoT is used to improve the motivation of students and keep them engaged and focused on the topic;
• IoT is a basis for laboratory exercises that supplement STEM lessons and lectures—predominantly in computer science but also in an interdisciplinary fashion, which includes electrical engineering and other subjects;
• IoT improves students’ digital competencies and 21st-century skills like digital literacy, computational thinking, critical thinking, design thinking, problem-solving skills, and collaboration;
• IoT enables multi- and interdisciplinary approaches to teaching and learning, gives the possibility to construct scientific knowledge in different disciplines and allows students to develop essential skills when solving STEM-oriented tasks;
• IoT allows students to assess their aptitude for future STEM careers, e.g., by participating in collaborative tasks aligned with industrial application contexts.

The second search query (SQ2) examines the security aspects of IoT in the STEM context. Very few publications mention and briefly discuss the topic of IoT security (RQ3). The focus is predominantly on the security of the collected IoT data and the need for appropriate STEM curricula, courses, and pedagogical strategies. These should not only introduce students to security methods for IoT applications but should also develop their skills in creating secure algorithms, code, and systems.

In general, IoT security does not represent a major topic in its own right in the educational field, yet. However, it may be identified as a potential topic to include and emphasize in future improvements in educational courses. This topic requires thorough research, and the authors consider a dedicated investigation.

The presented study has some limitations. The search was restricted to the Scopus and WoS databases because they are considered impactful, have rich technical content, and are comprehensive. Only papers in English were taken into account for further examination. The queries were focused on the last ten years, since it was expected that publications from 2015 to 2025 would be influential and trendsetting. Two search queries with specific keywords related to IoT in STEM education were submitted to Scopus and WoS to obtain a meaningful and targeted dataset to be bibliometrically analyzed in line with the research goal. The manuscripts were analyzed only by their titles, authors’ keywords, and abstracts to outline the general picture and research tendencies. The full-text content of the most impactful papers was qualitatively analyzed. The selection criterion for paper impact was the citation count, i.e., the publications with nine or more citations were analyzed. Other selection criteria, such as review/survey papers, publications in the most relevant journals and scientific forums, etc., could also be used. The results from our qualitative analysis, however, appear to give a good representation of the overall picture, recent topics, and research trends regarding the applications of IoT in STEM education.

6. Conclusions

This paper explores the applications of the Internet of Things in STEM education. The extensive bibliometric review of related publications extracted from global scientific databases identified prevalent research trends and topics, including artificial intelligence, project-based learning, hands-on learning, Arduino, sensors, programming, and robotics. The detailed qualitative analysis of the most influential recent papers outlined several distinct roles that IoT plays in the STEM educational field. The multidisciplinarity, connection to the real world, and the bearing on engineering and computer science make IoT highly suitable for STEM education. The created conceptual summary of the roles of IoT in STEM education shows the positive influences on students, teachers, and pedagogical approaches. The adoption of IoT as part of various teaching approaches and educational contexts enables the transformation of the teaching practice and enhances the learning process. Thus, the latter becomes more adaptive, interactive, and engaging, which should pave the way for students to gain interdisciplinary knowledge in a faster and more intuitive manner and for educators to acquire additional qualifications.

The topic of IoT security in STEM education is barely considered in the extracted dataset. There are some promising practical approaches reported in the literature; however, special attention needs to be paid to IoT security, e.g., incorporating IoT security into the curricula of STEM disciplines. Students and teachers can fully benefit from these innovative approaches when clear frameworks are established, and comprehensive training is provided. The ultimate goal is to acquire skills essential for the development and maintenance of secure IoT applications.

As general implications from the conducted study, some future research directions may be outlined. The identified gaps in the integration of IoT in STEM educational contexts should be addressed, including security issues and complexities related to privacy concerns. More research needs to be done to better understand the diversity of IoT technology applications, so that they are easier for teachers and students to adopt in the educational process. IoT should be included in curricula and in STEM courses together with structured policies for comprehensive practical and methodological training of teachers. Professionally developed STEM courses with dedicated setups should be delivered to all learners and educators interested in the subject area. Although IoT-supported courses in STEM education are practically oriented, they still do not consider interdisciplinary and industry-related aspects to a sufficient extent. The integration of such courses into real-life application scenarios and scenarios related to the needs of the industry should be strengthened. This could also serve to improve the support of dual educational systems. The critical analysis of the studied publications suggests that STEM partnerships be established with industry experts and professionals, enabling companies to engage with both teachers and students in secondary school and university education to support curriculum development, the delivery of coursework, and direct student interactions, thereby ensuring professional awareness and motivation for 21st-century careers.

IoT devices and setups are still relatively difficult to afford, although many open-source solutions are available. Government policies for investing in the development of freely available and remotely accessible diverse IoT applications related to various scientific topics can provide new opportunities for educational innovation. Further, there is a need for a methodology for the integration of various IoT setups to support constructive learning approaches through experiment-based, hands-on, and project-based activities, thus helping students understand complex scientific concepts.