JOURNAL OF LATEX CLASS FILES, VOL. 14, NO.

8, AUGUST 2021 1

Empowering STEAM Activities with Artificial

Intelligence and Open Hardware:
the BitDogLab
Fabiano Fruett, Fernanda Pereira Barbosa, Samuel Cardoso Z. Fraga, Pedro Ivo Aragão Guimarães

Abstract—Open-source hardware and software platforms have cost for consumers [3]. Furthermore, elementary and high
played an important role in democratizing access to technology school teachers face difficulties in using electronic components
and education in computer science and engineering. They have connected to a breadboard [4]. These difficulties mainly arise
significantly reduced costs and lowered barriers to entry for
enthusiasts, educators, and students to start working on electron- from common problems of poor contact on breadboards and
ics, programming and other technologies. Recent advancements the fragility of connections, which can lead to errors and
in open-source tools such as KiCad, MicroPython and the frustration during practical teaching activities. Ideally, elec-
Thonny development environment have the potential to accelerate tronic boards designed for educational purposes should be
low-budget educational applications, providing a smooth and easy to use and operate under the principle of open-source
consistent learning curve for users. Using these open-source
platforms, we designed and developed the BitDogLab, a tool for hardware, providing their manufacturing ”recipe” to encourage
STEAM activities development using embedded systems based on sharing and continuous improvement. It is also important
the Raspberry Pi Pico board, with uncomplicated programming for them to be well-documented to facilitate independent
supported by artificial intelligence through the GPT model. learning for students [5]. The programming environment of the
BitDogLab is an open-source hardware solution with a flexible board should prioritize logical thinking over the syntax of the
design that allows for continuous expansion for future develop-
ments in both software and hardware areas. Being based on open- programming language, as the latter can become an obstacle
source code, the functionality of BitDogLab can be enhanced by for students who are starting to develop their programming
integrating it with a variety of software resources. This includes, and computational thinking skills [6].
for example, programming the board to interact with databases, a An initiative evaluated the effectiveness of Fab Lab-based
functionality that can be crucial in applications involving artificial learning and shows that students are more interested in science
intelligence. Additionally, BitDogLab can be adapted to work in
conjunction with more sophisticated hardware, such as robotic lessons after using Fab Lab-based learning [7]. This research
systems or advanced sensors, expanding its range of educational corroborates an analysis of 225 studies comparing student
applications. The project’s open distribution also encourages performance in STEM courses using traditional learning ver-
community contributions, promoting a culture of collaborative sus active learning methods, revealing a significant positive
learning and innovation. effect of active learning on student performance [8]. Given
Index Terms—Open Hardware, STEAM Education, Artificial this scenario, it is essential to seek low-cost solutions that meet
Intelligence in Education, Project-Based Learning the requirements while providing quality learning experiences
and effective development of students’ skills. Therefore, we
I. I NTRODUCTION have developed BitDogLab: an open-source, low-cost, and
user-friendly hardware board designed specifically to teach
B RAZIL, along with other developing countries, faces a
significant challenge regarding the inclusion of technol-
ogy in elementary and secondary school classrooms, especially
STEAM concepts in an interactive and engaging manner. This
article will describe the development and capabilities of the
in activities or projects involving Science, Technology, Engi- BitDogLab, demonstrating how it can be programmed with
neering, Arts, and Mathematics (STEAM) [1]. This challenge the support of artificial intelligence to facilitate and expedite
is influenced by various factors such as teacher training, IT the learning process.
infrastructure, social and economic inequality, and more [2].
In Brazil, one particular factor is the reality of cascading II. D ESIGN OF THE B IT D OG L AB P RINTED C IRCUIT
taxes (federal, state, and municipal) that make technologi- B OARD
cally dependent solutions, subject to importation, prohibitively
expensive for many schools, especially public ones. These While the use of protoboards is a common educational tool,
taxes result in an approximately 120% increase in the final it can pose technical challenges for elementary and high school
teachers and students. Therefore, we consider it advisable to
We would like to express our profound gratitude to the IEEE Electron avoid the protoboard as the first form of contact with electronic
Device Society (EDS) for the generous grant that enabled us to develop STEM
activities. Our appreciation also extends to INCT Namitec (National Institute components. In this way, our project is based on a specially
of Science and Technology on Nano and Microelectronics). This initiative, designed printed circuit board (PCB) for educational activities.
conducted by CNPq (National Council for Scientific and Technological We have named the board BitDogLab, in honor of our mascot
Development) in Brazil, provided valuable resources and expertise in the
development of semiconductor devices that significantly contributed to the BitDog, which is a playful element of the Escola 4.0 project
success of our project. [9] to which this initiative belongs. This PCB uses as its
JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 2

(a) (b)
Fig. 1. Layout design of the BitDogLab PCB: (a) front and (b) back side.

central component a Raspberry Pi Pico, a module with a

low-cost (US$6) and high-performance ARM microcontroller,
which offers programming flexibility and robust features for
teaching concepts of Science, Technology, Engineering, Arts,
and Mathematics (STEAM) [10]. The BitDobLab PCB was
designed to have two parts that can be easily separated. Its
layout is shown in Figure 1a and Figure 1b. The upper part,
which has several drilled holes, is a terminal bar to facilitate
connection with alligator-type clamps. The main core of the
board is where the components are soldered. The main core of
the board measures 98 mm in length, 88 mm in width, and 1.6
Fig. 2. Detail of the choice of the type of component and how to solder the
mm in thickness. The terminal bar is 23 mm in length, while colored LED on the BitDogLab PCB. The choice depends on the user’s skills
retaining the same width and thickness as the main core. and knowledge, it can be: through-hole (PTH) or surface mount (SMD).
The BitDogLab PCB connects the microcontroller with
various peripheral components that can be soldered onto the
board as pedagogical needs demand or as the user’s interest depends on the quantity requested. For less than a dozen units,
increases. These connections are made with two levels of the approximate cost is $1 per board. The components of the
metallization, one on each face of the board. The board was project are detailed in the next section.
designed to offer flexible and accessible alternatives for elec-
tronic components, both in the through-hole assembly (PTH)
mode and in the surface-mount device (SMD) mode. This
approach allows the user to choose the best assembly strategy A. Materials
according to their skills and learning goals. For example, PTH
In addition to the BitDogLab board, the materials used in
components may be more suitable for beginners due to the
this project are: a five by five matrix of colored (RGB) LEDs,
ease of soldering, while SMD components can offer a more
two buzzers, three push buttons, a microphone, a joystick, an
advanced experience, closer to what is found in the industry.
RGB LED, a mini display, and an Insulation-Displacement
This feature is shown in Figure 2.
Connector (IDC) that allows externally connecting other com-
The repository with all Gerber files for manufacturing and ponents or boards, expanding its functions. In addition to
also the KiCad source file for BitDogLab is available at components such as resistors and capacitors. All of these
[11]. There are several manufacturers around the world that components are commercially available and easy to acquire.
offer PCB prototyping. The manufacturing cost of this board The list of materials is shown in Table I.
JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 3

TABLE I
B IT D OG L AB´S BILL OF MATERIALS

item tag description quantity

1 20-pin female header 2
2 Raspberry pi pico (H or W) 1
3 LED 5mm RGB LED common cathode 1
4 Rb 100 ohm resistor 1/8 W 5% 1
5 Rg, Rr 200 ohm resistor 1/8 W 5% 2
6 A, B, Reset button 3
7 A, B passive buzzer 2
8 Q1, Q2 NPN transistor BC337 2
9 R4, R5 330 ohm resistor 1/8W 5 % 2
10 M 5 x 5 Matrix RGB LEDs 5050 WS2812B 1
11 JOY KY023 analog joystick 1
12 reg Voltage regulator 3V3 LM1117 1
13 MIC PDM microphone module MP34DT01 1
14 C1-C6 100uF 16V electrolytic capacitor 6
15 DSP 0.96” I2C 128 x 64 OLED display 1
16 bat rechargeable battery holder for PCB 1
17 CHM charge module tp4056 mini USB 1

Fig. 4. Close-up view of the connection terminals of the BitDogLab that can
be connected with alligator clips.

an electrical connector that allows connections to be made with

insulated cables without the need to remove the insulation.
In practice, this means that electronic components can be
easily connected to cables, without the need for complex cable
preparation. This simplifies the assembly and maintenance
process, making it ideal for educational applications where
ease of use and expansibility are fundamental.

III. T HE PROGRAMMING ENVIRONMENT

The hardware we described earlier is the physical imple-
Fig. 3. Front view of the assembled BitDogLab with all its components. mentation of our tool for empowering STEAM Activities. The
core of this tool is a microcontroller that must be programmed
to command or receive signals from components, sensors,
B. Method for Hardware Implementation and peripherals. This microcontroller can be programmed in a
variety of ways [12]. However, the main contribution of this
With the BitDogLab PCB in hand, we suggest that each
article is to propose a method that allows anyone, even without
component be included, by soldering it onto the board, in the
any prior programming experience, to use this tool in STEAM
order that appears in Table I. In this way, after incorporating
projects in the classroom.
items 1 and 2, the user has the possibility to progress and
The programming environment comprises the Integrated
program using the approach presented in this article. As more
Development Environment (IDE), Thonny [13], and a Large
components are included, your options for use in STEAM
Language Model (LLM) of the GPT type. Thonny is a
activities increase along with your programming options. The
Micropython IDE that is incredibly beginner-friendly, with
soldering of the components onto the board is a task that
features like step-by-step debugging and variable monitoring.
requires appropriate guidance and safety precautions. With
The main interface of Thonny is shown in Figure 5 [13].
proper supervision, young people from 12 years old are capa-
Micropython is a compact and efficient implementation of
ble of performing quality soldering work. There are numerous
the Python language that can be used to program microcon-
tutorials and online videos that offer step-by-step instructions
trollers [14]. Chat GPT is an artificial intelligence feature [15]
on how to solder correctly. Although this task requires some
that can assist users in programming the BitDogLab based on
practice, once acquired, soldering skills become a lifelong
its Hardware Configuration Database.
useful competency. Furthermore, with the right guidance, the
activity can become not only educational but also fun. The
PCB with all the soldered components is shown in Figure 3. IV. H ARDWARE C ONFIGURATION DATABASE
The board has some terminals for connecting alligator clips. The Hardware Configuration Database (HCD), which we
In this way, the board’s expandability is facilitated by these propose, is a structured textual list containing information
terminals and also by a Insulation-Displacement Connector about the BitDogLab board’s hardware configuration. This
(IDC) with 14 vias, which are detailed in Figure 4. An IDC is database includes details about the arrangement and function
JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 4

pins is used for hardware expansion and is connected to

the Raspberry Pi Pico as follows: pin 1 with GND, pin 2
with 5V, pin 3 with 3V3, pin 4 with GPIO10, pin 5 with
GPIO28, pin 6 with GPIO9, pin 7 with analog GND. Pin
8 with GPIO8, pin 9 with GPIO17, pin 10 with GND,
pin 11 with GPIO16, pin 12 with GPIO19, pin 13 with
GND, pin 14 with GPIO18.
• A terminal bar that allows the connection of alligator clip-
type terminals is connected to the Raspberry Pi Pico as
follows: the DIG 0, 1, 2, and 3 electrodes are connected
to GPIOs 3, 2, 1, and 0, respectively. Additionally, this
terminal bar has five more electrodes connected to: analog
GND, GPIO 28, GND, 3V3, and 5V of the Raspberry Pi
Pico.
Fig. 5. Thonny’s Integrated Development Environment (IDE).
This Hardware Configuration Database is the key to making
the BitDogLab board an accessible and versatile tool for
STEAM education, providing all the necessary information for
of the hardware components, including but not limited to:
users and AI models to interact efficiently with the hardware.
connected peripherals and the GPIO pins of the Raspberry
Pi Pico microcontroller. This set of information simplifies
understanding the board’s architecture and allows users, and V. AI-A SSISTED P ROGRAMMING
even Artificial Intelligence models like ChatGPT, to interact In the following, we will describe how an artificial intelli-
efficiently with the hardware. For each peripheral, the database gence algorithm can be used as a Programming Support Tool
includes details about the connected GPIO pin, the expected for BitDogLab. When we say ”programming using Artificial
function of this component, and any specific parameters asso- Intelligence (AI)”, this may be interpreted as if the AI is
ciated with its use. The HCD not only serves as a reference executing the programming autonomously. However, current
guide but also aids in the development of STEAM (Science, AI models, including the GPT model, do not yet have this
Technology, Engineering, Arts, and Mathematics) applications capability. They cannot write complex computer programs
and activities with the BitDog PicoLab board. Moreover, it is independently, and more specifically, they do not have the
essential for customizing the AI model’s responses, enabling ability to program hardware boards. What AI models like GPT
it to provide specific and relevant information, particularly can do is assist humans in the task of programming. They
concerning programming code, based on the board’s hard- can provide code suggestions, explain programming concepts,
ware configuration. The BitDogLab Hardware Configuration assist in debugging errors, generate small snippets of code
Database is the textual list shown below: in response to specific requests, and so forth. In the case
• A common-cathode RGB LED has the red electrode of the BitDogLab project, from the Hardware Configuration
connected to GPIO 12 through a 200-ohm resistor, the Database, the AI can help transcribe and explain the code
green pin connected to GPIO 13 through a 200-ohm to control different components on the board, such as the
resistor, and the blue pin to GPIO 11 through a 100-ohm RGB LEDs, the buzzers, or the display from commands of
resistor. buttons, joysticks or sensors. Therefore, when we say that
• A button, identified as Button A, is connected to the BitDogLab board is ”programmed using AI”, this means that
Raspberry Pi Pico’s GPIO5. The other terminal of the AI is being used as a tool to assist humans in programming the
button is connected to the board’s GND. board. In this case, AI is used as an intelligent assistant that
• Another button, identified as Button B, is connected to can suggest, guide, and explain, but the final task of writing,
the Raspberry Pi Pico’s GPIO6. The other terminal of the testing, debugging, and implementing the code is guided and
button is also connected to the board’s GND. developed by humans.
• A buzzer, identified as Buzzer A, is connected to the
Raspberry Pi Pico’s GPIO21.
• Another buzzer, identified as Buzzer B, is connected to A. Method for hardware programming using AI
the Raspberry Pi Pico’s GPIO4. We propose a programming methodology, with AI support,
• The input pin of a 5-line by 5-column 5050 RGB LED that can be utilized in the development of a variety of interdis-
matrix type WS2812B (Neopixel) is connected to GPIO7. ciplinary learning projects and activities. This methodology is
• A KY023 type analog joystick has its VRx output con- characterized by an iterative approach, focused on the use of
nected to GPIO26 and its VRy output to GPIO27. Its SW a Hardware Configuration Database of open-source hardware,
button is connected to GPIO22. which is incorporated into the conversational platform of
• A 0.96-inch 128-column by 64-line OLED display with the AI model. Before starting to use the methodology, it is
I2C communication has its SDA pin connected to assumed that the user is already engaged in a STEAM project
GPIO14 and the SCL pin to GPIO15. that may involve electronics, sensors, actuators, and program-
• An Insulation-Displacement Connector (ICD) box of 14 ming in its implementation. The methodology is summarized
JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 5

Fig. 7. Front and back view of the PCB showing the basic components to
be soldered in the positions indicated on the board to carry out one of the
STEAM activities.

language syntax, the user can develop skills related to logic

Fig. 6. Flow diagram of the programming methodology with AI assistance. and computational thinking. As their interest increases, the
board can be expanded with new on-board components as well
as off-board ones.
in the flow diagram shown in Figure 6. The steps of the
methodology are detailed as follows: VI. STEAM ACTIVITY E XAMPLE WITH B IT D OG L AB
1) Loading the Hardware Configuration Database into In this section, we show an example of a STEAM activity
Chat GPT: This involves inserting the detailed informa- using the BitDogLab, which can be programmed using the
tion, in the form of a textual list, of the BitDogLab board proposed methodology. As an initial application example, we
into the GPT conversational platform. This includes in- prepared the PCB by soldering a few components, such as: two
formation about the layout of the hardware components, 20-pin rows (to connect the Raspberry Pi Pico), a single RGB-
the functions of each GPIO pin, etc. colored LED, and three resistors on the BitDogLab board.
2) Use of a Textual Flowchart: Editing or writing a These components are identified in Table 1 by their tags: LED,
Textual Flowchart: The user’s ideas for employing the Rr, Rg, and Rb. They are also identified on the PCB in the
BitDogLab in the STEAM project should be transcribed position they should be soldered, as shown in Figure 7.
in the form of a textual flowchart, using logical thinking. A pedagogical activity example, using only these compo-
This flowchart serves as a guide for the AI model, nents, involves programming so that the LED displays the
directing it on how to utilize the Hardware Configuration seven colors of the rainbow. This activity, although simple,
Database to assist in the creation of MicroPython code. serves as a programming exercise, and also as a multidisci-
3) Copying and Loading the Code into Thonny: After plinary teaching tool. In this programming exercise, one can
the code is created in collaboration with the AI, it is explore mathematical concepts by involving the calculation
copied and loaded into the Thonny development environ- of the percentages that determine the intensity of each color
ment, which is specific for MicroPython programming displayed by the LED. The same activity can also be used
and supports the Raspberry Pi Pico. to introduce science and art concepts related to color theory,
4) Download the Code to the Raspberry Pi Pico: The including primary colors and their blends. Thus, the activity
code is then downloaded to the Raspberry Pi Pico board incorporates STEAM disciplines concepts in an integrated
through the Thonny IDE. way. After soldering these few components, we follow the
5) Execution of the Program: From a command given by methodology previously presented. After opening the free
the user in Thonny, the Raspberry Pi Pico board executes GPT Chat interface, we load the Hardware Configuration
the loaded code, performing the programmed actions. Database (HDC) into the dialog window. In the same GPT,
6) Observation of Results: The results of the execution using the next dialog window, we load the textual flowchart
are observed by the user who analyzes and contrasts of operations to be performed. As an example, we can write:
them with their expectations and needs. At this point, Considering our Hardware Configuration Database, could
opportunities to increase both the hardware, by soldering you make a code to vary the colors of the RGB LED with
or connecting a new component, as well as modifying the sequence of the seven colors of the rainbow, starting
the code, can also be identified. with red and going to violet? The transition of the seven
7) Corrections with the AI: Any deviation from the colors should take 21 seconds and restart again.
expected behavior or need for change is discussed with After a few seconds, the response followed by the com-
the AI model, which uses the Hardware Configuration mented code appears in the Chat GPT response window, as
Database to suggest possible corrections to the code. shown in Figure 8.
This approach allows the student to go through all the stages The code can be copied simply by clicking on the ‘Copy
of a programming cycle and obtain results quickly, driving the code’ icon and pasting it into the Thonny´s window [11] as
process. Even without knowing details of the programming shown in Figure 9. The step-by-step guide on how to configure
JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 6

Pico) is out of the scope of this article’s innovation and will

not be detailed. For users less familiar with the initial setup
of the Raspberry Pi Pico using the Thonny IDE, including
the installation of Micropython, we recommend the ”Getting
started with Raspberry Pi Pico” document from the Raspberry
Pi Foundation [16] After loading the code into the BitDogLab,
the program begins to execute the requested tasks. The result
is shown in Figure 10.
Should there be any errors before the desired outcome is
achieved, the error code will appear in the Thonny Shell
window. This error information can be copied into a GPT
Chat window that will rewrite the program, with the user
simply observing and repeating the process. Observing and
comparing the achieved result with the desired one, based on
the textual flowchart, is a very important stage of the process
for the advancement of students’ STEAM skills.
For example, note that the difference between orange and
yellow, as shown in Figure 10, is very subtle and may cause
difficulty in distinguishing one from the other. This particular
result can become a challenge for students to improve the
code. Thus, they may be motivated to adjust the intensity
of each color in the code to deliver a result where yellow
is more distinguishable than orange. For this, they should
exercise concepts of proportion, rule of three, color blending,
and logical thinking. From this example, the user can expand
to include new possibilities, such as: playing the Buzzer at
a frequency proportional to the wavelength of each color’s
Fig. 8. Chat GPT model´s response to the structured question suggested by electromagnetic wave. This would be a suitable example to
the method of this article. explain the electromagnetic spectrum and relate its frequen-
cies to a practical outcome. Thus, in a more intuitive way,
students could understand the relationship between frequency
and wavelength in the electromagnetic spectrum.
Alternatively, they could include the display to show the
wavelength (in nm) of each color being displayed by the
colored LED. The commands to change the color could also
come from the joystick or from claps that would be picked up
by the microphone.

VII. D ISCUSSION AND C ONCLUSIONS

This work highlights the importance of an integrated and
suitable approach to the development of STEAM projects,
promoting a holistic and practical understanding, as well as
including hands-on experience with electronics and program-
ming. The experimental nature of the BitDogLab enables
students to apply theoretical knowledge, by creating and pro-
gramming real devices, thus solidifying meaningful learning.
The BitDogLab also offers interactive features that enhance
the educational experience. To start, we highlight an example
where the correlation of buzzer frequencies with electromag-
netic wavelengths emitted by the LED provides a practical and
multimodal approach to teaching complex concepts, such as
the electromagnetic spectrum. The implementation of various
user interfaces, involving: lights, sounds, and movement, in-
Fig. 9. Thonny IDE with code generated by Chat GPT. creases the interactivity of the learning experience provided by
BitDogLab and demonstrates its flexibility. Future research can
assess the effectiveness of this methodology in stimulating in-
the Thonny IDE along with the microcontroller (Raspberry Pi terdisciplinary learning and the development of programming
JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 7

(a) (b) (c) (d) (e) (f) (g)

Fig. 10. Result of the RGB LED displaying the seven colors of the rainbow in sequence.

skills. The application of this methodology in other contexts the National Academy of Sciences, vol. 111, no. 23, pp. 8410–8415,
and with different types of hardware is also a promising area 2014. [Online]. Available: https://www.pnas.org/content/111/23/8410
[9] “Inspire seus alunos com steam,” https://escola4pontozero.fee.unicamp.
for future investigation. Therefore, the central contribution of br/, accessed: 2023-06-03.
this work to the field of science and technology education [10] “Buy a raspberry pi pico,” https://www.raspberrypi.com/products/
is the proposal of an iterative programming process that raspberry-pi-pico/, accessed: 2023-06-03.
[11] “Bitdoglab repository,” https://github.com/Fruett/BitDogLab, accessed:
integrates AI into the development process, using the Hard- 2023-06-03.
ware Configuration Database (HCD). This approach, which [12] “Raspberry pi os,” https://www.raspberrypi.com/software/, accessed:
harmoniously unites artificial intelligence and open hardware 2023-06-03.
[13] A. Annamaa, “Thonny,: A python ide for learning programming,”
in the form of BitDogLab, has been specifically designed to in Proceedings of the 2015 ACM Conference on Innovation and
simplify the programming task for both students and educators. Technology in Computer Science Education, ser. ITiCSE ’15. New
In doing so, it not only demystifies a complex subject, such York, NY, USA: Association for Computing Machinery, 2015, p. 343.
[Online]. Available: https://doi.org/10.1145/2729094.2754849
as dealing with complex coding syntax, but also actively [14] “Micropython,” https://micropython.org/, accessed: 2023-06-03.
encourages the development of critical thinking and problem- [15] OpenAI, “Gpt-4 technical report,” 2023.
solving skills. In this way, BitDogLab can be an empowering [16] “Getting started with raspberry pi pico,” https://projects.raspberrypi.org/
en/projects/getting-started-with-the-pico/2, accessed: 2023-06-03.
tool for STEAM activities, providing an environment where
curiosity is rewarded with knowledge and challenges serve as
stepping stones to understanding. This confluence of elements
propels BitDogLab beyond a mere educational tool, shaping
it as a catalyst to inspire lifelong learners in the ever-evolving
landscape of STEAM education.

R EFERENCES
[1] S. Marginson, R. Tytler, B. Freeman, and K. Roberts,
“STEM: country comparisons: international comparisons of
science, technology, engineering and mathematics (STEM)
education. Final report.” 1 2013. [Online]. Available:
https://dro.deakin.edu.au/articles/book/STEM country comparisons
international comparisons of science technology engineering and
mathematics STEM education Final report /20951737
[2] S. Mejias, N. Thompson, R. M. Sedas, M. Rosin, E. Soep,
K. Peppler, J. Roche, J. Wong, M. Hurley, P. Bell, and B. Bevan,
“The trouble with steam and why we use it anyway,” Science
Education, vol. 105, no. 2, pp. 209–231, 2021. [Online]. Available:
https://onlinelibrary.wiley.com/doi/abs/10.1002/sce.21605
[3] “Brazilian Tax aspects on import of services,” https:
//www.pnst.com.br/en/articles/tax-consulting-and-litigation/
brazilian-tax-aspects-import-services/, accessed: 2023-06-03.
[4] K. Peppler, M. Sedas, and N. Thompson, “Paper circuits vs. breadboards:
Materializing learners’ powerful ideas around circuitry and layout de-
sign,” Journal of Science Education and Technology, pp. 1–24, 05 2023.
[5] R. Heradio, J. Chacon, H. Vargas, D. Galan, J. Saenz, L. De La Torre,
and S. Dormido, “Open-source hardware in education: A systematic
mapping study,” IEEE Access, vol. 6, pp. 72 094–72 103, 2018.
[6] P. Trivedi, P. Kajgaonkar, A. Kulkarni, N. Kolte, and B. Kanawade,
“System model for syntax free coding,” in 2019 Global Conference for
Advancement in Technology (GCAT), 2019, pp. 1–5.
[7] M. A. Togou, C. Lorenzo, G. Cornetta, and G.-M. Muntean, “Assessing
the effectiveness of using fab lab-based learning in schools on k–12
students’ attitude toward steam,” IEEE Transactions on Education,
vol. 63, no. 1, pp. 56–62, 2020.
[8] S. Freeman, S. L. Eddy, M. McDonough, M. K. Smith, N. Okoroafor,
H. Jordt, and M. P. Wenderoth, “Active learning increases student
performance in science, engineering, and mathematics,” Proceedings of
JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 8

Fabiano Fruett Fabiano Fruett received his degree

in electrical engineering in 1994 from the Uni-
versidade Estadual Paulista - UNESP-Ilha Solteira,
and his Master’s degree in microelectronics in 1997
from the Universidade Estadual de Campinas - UNI-
CAMP. In the same year, he began his work as
a researcher in the field of microelectronic sensors
at the Delft University of Technology, Netherlands.
Fruett obtained his Ph.D. in September 2001, where
the primary findings of his doctoral research were
published in over a dozen scientific articles and a
book authored by him. Since 2002, he has been working as a faculty member
at the School of Electrical and Computer Engineering (FEEC) at UNICAMP.
He is currently an Associate Professor III in the renamed Department of
Electronics and Biomedical Engineering of the FEEC. His research interests
are in the areas of electronic instrumentation, Internet of Things (IoT), and
open-source STEAM learning tools.

Fernanda Pereira Barbosa Master’s degree in

Chemistry from the Department of Chemistry at
UFSCar. Supporter of the Escola 4.0 project. Curator
of social networks for the project Escola 4.0.

Samuel Cardoso Zampolli Fraga Undergraduate

student in Electrical Engineering at State University
of Campinas - Unicamp.

Pedro Ivo Aragão Guimarães Mr. Guimarães is

currently a M.Sc. student in Electronic Engineering
at the Federal Institute of Education, Science and
Technology of Paraı́ba—IFPB, Brazil. He received
the B.Sc. in Electronic Engineering from the Federal
University of Campina Grande-UFCG, Brazil, in
2021.