Apertura

The experiment to execute the challenge was conducted with two groups of 30 fifth-semester undergraduate students enrolled in a Mechanisms course. Students were organized into teams of three or four members. The first group (control group) only used product design tools to solve the challenge. The second group (experimental group) used product design tools in combination with ChatGPT to solve the challenge. In both groups, the students did not have prior knowledge of the product design tools or their usefulness before the experiment. To ensure a fair comparison between the two groups, the teacher used an evaluation rubric focused on the effectiveness of problem-solving through the application of product design tools, independent of the use of AI tools.

Initially, the professor explained the theoretical concepts and examples of the product design tools in both groups:

The House of Quality emphasizes the importance of prioritizing the elements of the project and establishing comparisons against competitors.

The selection of product design tools for this study was guided by their proven effectiveness in engineering education and professional design processes. The House of Quality was chosen for its ability to translate customer or project requirements into prioritized design elements, fostering a structured approach to addressing user needs while benchmarking against competitors. The Function Block Diagram was included due to its utility in systematically mapping the functional requirements of a design, providing students with a clear understanding of the relationships between subsystems and their roles in achieving the overall objective. Finally, the Failure Mode and Effects Analysis (FMEA) was employed to help students identify potential risks, evaluate their impact, and prioritize mitigation actions and critical skills to ensure robust and reliable designs. Together, these tools were selected for their complementary strengths in guiding the ideation, analysis, and refinement stages of the design process, making them particularly suitable for the context of the robotics challenge.

In addition to the class, a series of videos explaining the product design tools was prepared and provided to the students to reinforce their learning and understanding. An evaluation was conducted to ensure that the students had reviewed the material.

In the case of the experimental group, the professor introduced and explained the use of ChatGPT. Students participated in a session where the application of ChatGPT was demonstrated alongside the product design tools to help them understand how this AI tool could be utilized to solve different problems. For instance, the professor asked ChatGPT about potential design failures that might arise in a project (e.g., flat knitting machine automation) with characteristics similar to their challenge, as illustrated in Figure 1. Following this explanation and demonstration, the professor assigned exercises for students to practice using ChatGPT. These exercises ensured that students could effectively formulate prompts for the AI tool and critically analyze the responses to confirm their accuracy and feasibility in addressing the problem.

Figure 1

The challenge’s objective was to design and build a mechanism mounted on a mobile robot capable of removing a flag from another mobile robot. In the next class sessions, the professor explained the robot’s characteristics and guided the initial ideation process. The type of mobile robot selected for the mechanism was a Sumo Robot, as shown in Figure 2.

Figure 2

Both groups had the same amount of time to complete the challenge and evaluate the results. The experimental group was told that they could use ChatGPT at any stage of the robot design and implementation process. When the students needed help, they asked the teacher to check the elaboration of the prompts. Although a deadline was set for the delivery of the project, the professor in both groups reviewed the progress made and answered any doubts that the students might have.

CBL provides a framework for engaging students in real-world problem-solving by encouraging interdisciplinary collaboration and iterative thinking. In the context of mechatronics, CBL aligns well with the need to integrate theoretical knowledge with practical applications, offering a dynamic platform for students to develop technical and non-technical skills simultaneously. The first task assigned to students was to prepare a report detailing some of the results of their design. Figure 3 presents some sketches from the groups. Although the initial sketches do not demonstrate a significant difference between the two groups, it is essential to note that the students in the experimental group were the first to deliver the requested progress.

Figure 3

As previously mentioned, the students in the experimental group also used ChatGPT, and Figure 4 shows a prompt they used; the responses obtained formed the basis for the idea-generation process. ChatGPT was integrated as a supportive tool throughout the CBL framework. During the ideation phase, it facilitated brainstorming by generating diverse perspectives and potential solutions. In the design phase, ChatGPT provided feedback on technical specifications and helped refine project ideas. The tool helped students troubleshoot and optimize their designs during implementation by explaining complex concepts and suggesting improvements. Finally, in the evaluation phase, ChatGPT supported students in articulating their findings and reflecting on their learning processes.

Figure 4

In summary, to ensure replicability, a detailed implementation protocol was followed. The experiment consisted of the following steps:

1) Training on Product Design Tools: Students in both groups attended an introductory session where the House of Quality, Function Block Diagram, and FMEA were explained with practical examples. Supplementary video materials were provided, and a follow-up quiz was conducted to ensure understanding.

The following resources were required for the implementation:

Technological Resources: Laptops with internet access for running ChatGPT, and software for design tools. • Physical Materials: Motors, sensors, microcontrollers, batteries, and other components for building the robots.

Three questionnaires were administered to the students in the group that utilized the product design tools and ChatGPT to address the challenge. The first questionnaire evaluated their prior use of ChatGPT in projects for other courses. As in shown in Figure 5, nearly half of the students had already used ChatGPT for academic projects before this experiment.

Figure 5

In that questionnaire, students who had used ChatGPT in their projects made comments such as “(I use it) when I feel out of ideas” or “It helped me to generate new ideas.”

The second questionnaire was distributed after the videos were provided to the students. It included the following questions for each tool:

a) How useful do you consider the House of Quality for your mechatronic projects?

A Likert scale ranging from 1 to 5 was used, with 1 representing the minimum value and 5 the maximum. The results in Figure 6 show that after introducing the tools, students were optimistic about using the product design tools in their projects. Although the results are presented for the House of Quality, a similar opinion was expressed regarding the Function Block Diagram and the FMEA.

Figure 6

Finally, another questionnaire was administered to the students in the experimental groups after the challenge was completed, with the following questions:

1) What was the easiest part of this project/ challenge for you?

A Likert scale ranging from 1 to 5 was used, with 1 representing the minimum value (very little) and 5 the maximum value (very much). The results for the first two questions are presented in Figure 7. Notably, half of the students reported that the ideation was the easiest part, likely because the provided tools facilitated the quick generation of design ideas. In contrast, the most challenging aspect was the implementation, as is often the case in projects involving the construction of physical devices.

Figure 7

Figure 8 presents the results for questions 3 and 4. On average, students acknowledged the usefulness of the provided tools and expressed a greater likelihood of using ChatGPT for future projects.

Figure 8

Finally, for those who indicated they would use ChatGPT for future projects, Figure 9 shows their preferences regarding the tools they would likely use in conjunction with ChatGPT.

Figure 9

The integration of ChatGPT within the CBL framework enhances the students’ problemsolving process rather than a replacement for the foundational principles of CBL. By acting as a collaborative assistant, ChatGPT reinforces critical thinking and creativity, which are core objectives of CBL, particularly in tackling real-world challenges in mechatronics education.

The final evaluation focused on two mechanisms: one developed by the control group and the other by the experimental group, as illustrated in Figure 10. These mechanisms were assessed based on five criteria:

a) Performs the function adequately: This criterion evaluated the mechanism’s ability to effectively remove a flag from another robot, with an emphasis on precision and reliability during repeated trials.

Figure 10 Note: The requested mechanisms are inside the area of the yellow rectangles

Each criterion was rated on a Likert scale ranging from 1 to 10, where 1 corresponded to a marginal fulfillment of the criterion, and 10 indicated complete fulfillment. The evaluation process was designed to ensure objectivity and comprehensively highlight each mechanism’s strengths and weaknesses.

As shown in Figure 11, the mechanisms produced by the experimental group outperformed those from the control group across all criteria. The experimental group’s mechanism demonstrated superior functionality, repeatability, and ease of use. One likely reason for this difference is that the experimental group benefited from integrating ChatGPT during the design process. ChatGPT provided valuable insights and suggestions during the ideation phase, which may have contributed to a more refined and effective final design.

Figure 11

Moreover, the originality of the mechanism developed by the experimental group was particularly noteworthy, suggesting that the use of AI tools like ChatGPT can enhance creativity in the design process. The experimental group’s clear advantage in these evaluations highlights the potential benefits of integrating AI-driven idea generation with traditional product design methodologies in educational settings.

The results of the Mann-Whitney U Test revealed no statistically significant differences between the control group and the experimental group across all evaluated metrics. While these findings suggest that the integration of ChatGPT did not result in measurable distinctions in the specific scores analyzed, the instructor observed notable differences in the students’ creative processes and the quality of the solutions presented. These qualitative observations highlight the impact of ChatGPT as a tool for enhancing student engagement and innovation during the project.

The integration of ChatGPT within the CBL framework complements and enhances the students’ problem-solving abilities without replacing the foundational principles of CBL. Acting as a collaborative assistant, ChatGPT fosters critical thinking and creativity, aligning closely with the core objectives of CBL, particularly in addressing real-world challenges in mechatronics education. By supporting brainstorming, refining design ideas, and offering iterative feedback, ChatGPT contributes to a richer and more dynamic learning experience for students.

In general, it is important to highlight that the authors designed the questionnaires used in this study for the context of the robotics challenge. Their primary objective was to collect exploratory and direct information regarding students’ perceptions of the product design tools and ChatGPT integration within the challenge. Although the questionnaires were not formally validated, they consisted of straightforward, clear, and specific questions, ensuring consistent understanding among participants. This approach enabled an objective assessment of key aspects, such as the perceived usefulness of the tools and the challenges faced during the project.

The exploratory nature of these questionnaires aimed to provide preliminary insights that could serve as a foundation for hypothesis generation and guide future research. However, we acknowledge that formal validation of the instruments would enhance the reliability and generalizability of the findings. Therefore, in future studies, we plan to adopt more rigorous processes for designing and validating instruments, ensuring greater precision and validity in the collected data.

The findings of this study carry significant pedagogical implications, particularly in the context of integrating AI tools like ChatGPT into educational frameworks such as CBL. The ability of ChatGPT to enhance ideation and creativity underscores its potential as a supportive tool for fostering higher-order thinking skills among students. By streamlining the generation of ideas and providing diverse perspectives, ChatGPT can help learners move beyond rote memorization toward more critical and creative problem-solving approaches. However, its implementation must be guided by well-defined pedagogical strategies. Educators should design activities that leverage the tool’s capabilities and encourage students to critically evaluate its outputs, ensuring that AI-generated content complements rather than replaces student effort. This aligns with constructivist principles, where learning occurs through active engagement and reflection, emphasizing the role of the educator as a facilitator in guiding the meaningful integration of technology.

From an ethical perspective, the widespread adoption of ChatGPT in education raises important concerns. These include potential over-reliance on AI, the perpetuation of biases embedded in training data, and challenges in ensuring academic integrity. Institutions must establish clear policies and guidelines to address these issues, promoting responsible use of AI tools. For instance, incorporating discussions on ethical AI practices into the curriculum can empower students to recognize the limitations and potential biases of ChatGPT while fostering a critical mindset. Additionally, educators must ensure that using ChatGPT does not create inequities, such as favoring students with better access to technology. By addressing these ethical considerations, educational institutions can create an environment where ChatGPT is used responsibly to enhance learning outcomes without compromising the values of fairness, originality, and integrity.