1. Introduction
The construction industry is experiencing a profound transformation driven by rapid advancements in technology, particularly the increasing integration of artificial intelligence (AI) [
1]. This technological revolution is reshaping traditional practices across various construction phases, including project planning, design, site management, and safety. As these innovations continue to evolve, the roles and responsibilities of future construction professionals are undergoing significant changes. Construction management students, who represent the industry leaders of tomorrow, must develop the skills and knowledge necessary to navigate this shifting landscape. However, their perceptions of AI, its potential applications, and its implications for their future careers remain insufficiently explored [
2].
Generative Artificial Intelligence (GAI) stands out as a transformative technology, significantly impacting the educational landscape and sparking considerable attention and debate [
3,
4,
5,
6]. With the increasing adoption of GAI, students have enhanced their problem-solving abilities and assignment completion efficiency, often reducing reliance on instructors for guidance [
7]. However, this progress is accompanied by ethical challenges and academic integrity concerns, such as potential cheating and unequal access to software [
8,
9]. GAI applications span a wide range of domains, including speech, text, video, research, coding, and image generation.
The Architecture, Engineering, and Construction (AEC) industry, in particular, has witnessed substantial growth over the past decade, cementing its role as one of the most influential global markets. Employment in the U.S. construction sector steadily increased from 2011 to 2019 and demonstrated resilience during the COVID-19 pandemic, with unemployment impacts less severe than the overall average and employment levels rebounding to pre-pandemic figures.
Construction management graduates are regarded as future leaders in the implementation of artificial intelligence (AI) in construction due to their expected progression into managerial and decision-making roles. Their education and training provide them with expertise in project planning, resource allocation, and problem-solving, all of which are essential for integrating AI-driven technologies into the industry. As they advance into leadership positions, their perception of AI will play a crucial role in determining how quickly and effectively it is adopted [
10].
If construction management graduates recognize AI as a valuable tool for improving efficiency, reducing costs, and enhancing safety, they are more likely to advocate for its adoption. This may involve supporting investments in AI-powered scheduling, cost estimation, safety monitoring, and automation. However, if they perceive AI as a disruptive force that threatens traditional practices or job security, they may resist its integration, leading to slower adoption rates. Their perception will also influence company policies, investment strategies, and collaborations with AI developers, ultimately shaping the future of AI in construction [
11].
Understanding how undergraduate construction management students perceive AI is crucial for designing education and training programs that align with the industry’s technological advancements. The adoption of AI in construction presents both opportunities and challenges, requiring professionals to adapt to new roles, acquire advanced technical skills, and integrate AI into their workflows. Future construction leaders must not only understand emerging technologies but also harness them effectively to drive industry progress.
Artificial intelligence, broadly defined as the ability of computers to learn, make decisions, and perform tasks characteristic of human intelligence, draws from disciplines such as mathematics, computer science, biology, psychology, and cybernetics. AI enables the development of algorithms and programs capable of solving problems autonomously. Despite employing approximately 7.22% of the global workforce, the construction sector continues to struggle with low labor productivity, resulting in significant waste of labor, materials, and financial resources. Given its substantial economic impact, the industry requires effective management practices to enhance productivity. Estimates suggest that improving productivity by 50.11% to 60.04% could generate an additional RUB 1.6 trillion in annual profits, further contributing to global GDP [
12]. By fostering a positive perception of AI among construction management graduates, the industry can better position itself to leverage AI’s potential, ultimately driving innovation, efficiency, and profitability.
AI encompasses nine major subfields, including knowledge-based systems, computer vision, robotics, natural language processing, automated planning and scheduling, optimization, and machine learning [
2]. In educational contexts, AI primarily manifests through knowledge-based systems, automation, natural language processing, and adaptive learning—tools such as intelligent tutors, automated feedback systems, and grading mechanisms exemplify these applications [
13]. These subfields are already influencing engineering education and faculty practices, with the potential to drive further innovation.
As GAI technologies advance, educators face pressing challenges, including ethical considerations [
8,
9], data privacy [
14], and equitable access [
15]. Critical questions arise about whether faculty should actively teach AI tools, restrict their use to uphold academic integrity, or establish regulatory guidelines to govern their implementation [
6]. Addressing these questions will be pivotal in shaping the future role of AI in education and the construction industry.
4. Research Methods
This study utilized a quantitative analysis approach to analyze data collected from undergraduate students in the Department of Construction Management at a university located in the Southeastern United States. The data collection process incorporated a survey designed to capture both quantitative and qualitative insights, tailored to the nature of the questions.
The survey instrument was carefully developed and reviewed by the UNF Institutional Review Board (IRB) to ensure compliance with ethical research standards. Following IRB approval, UNF’s Center for Instruction and Research Technology (CIRT) assisted in designing the survey using Qualtrics and facilitated its anonymous distribution to students within the Construction Management program.
A total of 99 responses were obtained, representing a diverse sample of freshmen, sophomores, juniors, and seniors. The survey comprised 18 questions grouped into five primary themes.
Table 1 presents a detailed breakdown of the survey questions, categorized by theme and the corresponding analytical approach employed based on the question type.
For quantitative analysis, both descriptive statistics and inferential statistical methods were employed to explore differences in students’ perceptions of AI by year of study. Analysis of Variance (ANOVA) and descriptive data analysis were used to compare group means across different student groups. ANOVA was applied to questions where comparisons were made between more than two groups (e.g., differences among freshmen, sophomores, juniors, and seniors). This analysis allowed for the examination of statistical significance in the observed group differences, providing insights into whether variations could be attributed to chance.
5. Findings and Results
The findings are based on survey responses collected from 99 undergraduate students enrolled in the Department of Construction Management at a university located in the Southeastern United States. This analysis incorporates frequency counts, percentages, and Analysis of Variance (ANOVA) to explore variations in perceptions of AI across students’ academic years.
Table 2 provides a descriptive analysis of the respondents’ demographic and academic profiles. The majority of participants are male, accounting for 84% (84 students), while females represent 15% (15 students). This gender distribution aligns with trends in the Department of Construction Management and reflects broader industry patterns, where male representation has historically been dominant. A minor portion of data (1%) is missing due to one participant not specifying their gender.
The sample includes students from all academic levels, with distribution as follows: freshmen (2%, 2 students), sophomores (4%, 4 students), juniors (59%, 59 students), and seniors (35%, 35 students). Juniors form the largest group, likely reflecting program enrollment patterns, while freshmen and sophomores constitute smaller segments of the population. Additionally, 6% (6 students) of the respondents are veterans, and the remaining 94% (94 students) are non-veterans.
To address the first objective, we examined students’ awareness and understanding of AI applications in the construction industry.
Table 3 shows students’ familiarity and experience with AI in construction. Regarding familiarity with AI, students’ self-reported knowledge levels varied significantly:
15% (15 students) indicated they were “Not familiar” with AI;
60% (60 students) described themselves as being “Somewhat familiar”;
23% (23 students) considered themselves “Very familiar”.
These findings suggest that while the majority of students have at least a basic awareness of AI concepts, only a quarter feel highly knowledgeable. In terms of practical experience, 55% (55 students) reported having some level of experience with AI technologies, whereas 43% (43 students) indicated no experience. A small percentage (2%, two students) did not respond to this question.
This distribution underscores that over half of the respondents have had some exposure to AI, but there is a significant portion with limited or no practical experience. This suggests a need for greater integration of AI-related content into the Construction Management curriculum to ensure students are adequately prepared to engage with these technologies in professional settings.
The findings highlight disparities in students’ familiarity and experience with AI, which may impact their readiness to incorporate AI into construction processes. Given the increasing importance of AI in the construction industry, the results emphasize the need to enhance educational opportunities and practical exposure to AI within the program. Addressing these gaps could foster a workforce better equipped to leverage AI-driven innovations effectively.
Table 4 summarizes students’ perception of AI’s potential impact on job roles, productivity, safety, and innovation in the industry, addressing this study’s second objective. The mean score for Q7 is 3.77, indicating that, on average, students perceive AI as having a “Significant Impact” on the construction industry over the next 5–10 years. Similarly, the mean score for Q18 is 3.69, reflecting a generally positive outlook toward AI’s potential to create new opportunities, such as generating jobs and enhancing project efficiency. These relatively high mean scores, coupled with slight negative skewness, suggest a strong overall optimism among students about AI’s transformative potential and its future role in the industry.
In contrast, the mean score for Q9 is 2.66, which falls closer to the “Neutral” point. This indicates that students are somewhat undecided about AI’s ability to improve safety and reduce accidents on construction sites. The higher standard deviation and flatter kurtosis observed for Q9 suggest a broader range of opinions, reflecting diverse perspectives and potential ambivalence on this topic. This uncertainty may stem from a lack of familiarity with or exposure to specific examples of AI applications in construction site safety.
Students express enthusiasm about AI’s broader impact on the construction industry and its potential to drive innovation and create opportunities. However, the mixed perceptions regarding AI’s role in improving safety highlight an area for further educational focus. Incorporating real-world examples and case studies into the curriculum could help address this gap by demonstrating how AI technologies are being used to enhance safety on construction sites effectively. Providing concrete evidence of AI’s safety applications may not only clarify its benefits but also inspire greater confidence among future construction professionals.
Skewness and kurtosis are statistical measures that describe the shape and distribution of a dataset. Skewness indicates the degree of asymmetry in the data. A positive skew means more respondents gave lower ratings, while a negative skew suggests higher ratings were more common. For example, in our study, Question 7 (AI’s impact on the construction industry) shows a slightly negative skew, indicating that most respondents are optimistic about AI’s future role. However, an absolute skewness value below 1 suggests the data are approximately symmetric and not significantly skewed [
40].
Kurtosis measures the “tailenders” of a distribution or its tendency to produce extreme values. A high kurtosis indicates a sharper peak and heavier tails, meaning a greater likelihood of outliers. In our study, the kurtosis values for Question 18 suggest some variability in respondents’ views on AI’s ability to create new opportunities. While many see AI as beneficial for construction, some express concerns about job displacement or implementation challenges. However, an absolute kurtosis value below 3 suggests the data do not have excessive tails compared to a normal distribution [
41]. In this study, the highest absolute values of skewness and kurtosis are 0.436 and 0.555, respectively. These values indicate that the dataset is approximately symmetric and does not exhibit extreme tails, aligning closely with a normal distribution.
In addition, to validate the assumptions of normality and homogeneity of variance, the Shapiro–Wilk test and Levene’s test were employed. The statistics from the Shapiro–Wilk tests were close to 1, indicating that the variables followed a normal distribution and, thus, failed to reject the normality assumption. Moreover, Levene’s tests were conducted to check for homogeneity of variance before ANOVA was used. The results failed to reject the assumption of equal variance, indicating that the variances of the groups were approximately equal; therefore, we proceeded with ANOVA.
5.1. Perception Difference by Sex
Students’ awareness and understanding of AI applications in construction would differ by sex, school year, and veteran status. The following section presents the results.
Table 5 presents the results of the ANOVA test, which evaluates whether students’ awareness and understanding of AI applications in construction vary by gender. The test yielded a
p-value greater than 0.05, indicating that the null hypothesis could not be rejected. This finding suggests no statistically significant difference in AI perceptions between male and female students.
These results indicate that gender does not play a substantial role in shaping attitudes toward AI within this sample. Both male and female students demonstrate similar levels of awareness and understanding of AI applications in the construction industry. This consistency highlights a shared perspective among students, irrespective of gender, on the importance and potential impact of AI in their field.
5.2. Perception Differences by School Year
Table 6 presents the ANOVA test results assessing whether students’ awareness and understanding of AI applications in construction vary by academic standing. For ANOVA to produce meaningful results, each group must contain multiple cases to ensure sufficient within-group variance for comparison [
42]. In this study, only one freshman participant was represented, which led to their exclusion from the analysis to meet these requirements.
The ANOVA test yielded a p-value greater than 0.05, indicating that the null hypothesis could not be rejected. This finding suggests no statistically significant differences in AI perceptions across academic years. Students from different levels (e.g., sophomores, juniors, and seniors) exhibit similar attitudes and understanding of AI within this sample.
These results imply that academic standing does not substantially influence students’ perceptions of AI in the construction industry. This shared perspective across school years highlights a consistent level of awareness and understanding of AI applications, regardless of students’ progression within the program.
5.3. Perception Differences by Veteran Status
Table 7 presents the ANOVA test results examining whether students’ awareness and understanding of AI applications in construction differ by veteran status. The ANOVA test yielded a
p-value greater than 0.05, leading us to fail to reject the null hypothesis. This indicates no statistically significant difference in AI perceptions between veteran and non-veteran students. Thus, veteran status does not appear to significantly influence students’ attitudes toward AI, as both groups in this sample demonstrate similar views on the topic.
The ANOVA test results across gender, school year, and veteran status indicate no statistically significant differences in students’ awareness and understanding of AI applications in construction. P-values greater than 0.05 in each test led to failure to reject the null hypothesis, suggesting that factors such as gender, academic standing, and veteran status do not meaningfully influence students’ attitudes toward AI. Overall, students in this sample, regardless of these demographic variables, share similar perspectives on AI in construction.
5.4. AI Technologies Preferences in Construction (Q17)
The most popular responses were related to AI-powered design and modeling software, with 13% of respondents selecting this option. Drones for surveying and site monitoring were also a popular choice, selected by 13% of respondents. The combination of multiple technologies suggests that respondents see value in integrating different AI solutions. The chi-square test results comparing Q17 responses with academic levels: the
p-value of 0.2276 suggests that there is not a statistically significant relationship between academic level and technology preferences.
Figure 1 shows the distribution of preferences for different AI technologies in construction, with clear bars and percentage labels for easy interpretation.
The analysis of Q17 has shown that there was no significant relationship between academic level and preferences for AI technologies in construction, as indicated by the chi-square test. The visualization effectively highlights the distribution of preferences, with AI-powered design and modeling software being the most popular choice.
5.5. Interest on Course (Q10)
We assumed that students’ awareness and understanding of AI applications in construction reflect differences in educational interest.
Table 8 presents the results, with
p-values for Q7 and Q18 below 0.05, indicating significant differences in educational interest based on students’ perceptions of AI.
To identify the specific groups contributing to these differences, a Scheffé test was performed, with the results displayed in
Table 9. The table displays the results of the Scheffé test, categorizing students by their level of interest in AI applications in education: Group 1 reflects high interest; Group 2 has moderate interest, and Group 3 has no interest. In the “Differences” row, numbers in parentheses identify groups that are significantly different from the reference group (the group number outside the parentheses).
For Q7, the mean values rise as interest levels increase, indicating that students with stronger perceptions of AI are also more eager to learn about its applications in construction. Similarly, Q18 results reveal that students with a pronounced interest in educational opportunities score higher in AI perception. This suggests that students with more positive perceptions of AI are also more motivated to pursue educational pathways focused on AI applications, pointing to a potential link between positive AI perceptions and academic engagement in related fields.
5.6. Intention to Take Course (Q11)
This study also assesses the differences in students’ intentions to take the course based on their level of awareness and understanding of AI applications in construction.
Table 10 presents the results, with
p-values for Q7, Q9, and Q18 below 0.05, indicating significant differences in students’ intentions to enroll in AI-related courses within the field of construction based on their perceptions of AI.
To pinpoint the specific groups contributing to these differences, a Scheffé test was conducted, and the results are displayed in
Table 11. The table presents the Scheffé test results, categorizing students by their level of intention to enroll in AI courses within construction engineering: Group 1 represents those with no intention, Group 2 with moderate intention, and Group 3 with high intention. Except for Q7, no significant differences were observed among groups based on intention level.
For Q7, students with a high intention to pursue educational opportunities showed higher perception scores. This suggests that students with more positive perceptions of AI are more inclined to engage in educational pathways focused on AI applications in construction.
5.7. Necessity of AI Education (Q12)
We posited that students might differ in their perceptions of the necessity of AI education based on their views on AI.
Table 12 shows that only the
p-value for Q7 is below 0.05, indicating significant differences in students’ perceptions of the necessity of AI education based on their views on AI.
To identify the specific groups contributing to these differences, a Scheffé test was conducted, and the results are displayed in
Table 13. The table displays the Scheffé test results. Students who perceive AI education as necessary demonstrated higher perception scores. This suggests that students with stronger perceptions of AI are more likely to consider learning AI applications to be important.
6. Conclusions and Discussions
The findings of this study emphasize the critical need to reshape education and training for construction professionals in response to the expanding influence of AI in the industry. There is a growing demand for programs that combine construction expertise with AI proficiency, and universities are starting to address this need by integrating AI-focused curricula. Preparing students for an increasingly technology-driven construction industry will be essential to ensuring that the future workforce can thrive in an AI-enhanced environment. With AI rapidly reshaping job roles and opening new opportunities, individuals who are both adaptable and technologically skilled will be best positioned to succeed.
AI applications are already revolutionizing key areas of construction, including design, project management, site safety, automation, and facility management. The integration of AI allows construction companies to optimize workflows, reduce costs, enhance safety measures, and elevate the quality of their projects. With ongoing advancements in AI research, the technology’s potential to transform the construction industry continues to expand.
Among undergraduate students, however, there is a varied understanding of AI, with notable disparities in AI literacy across different academic disciplines. Students in technical fields generally demonstrate a stronger grasp of AI concepts, while those in non-technical fields may lack a deeper understanding despite recognizing AI’s relevance to their future careers. Barriers such as limited access to AI coursework, the technical complexity of AI, and common misconceptions impede students’ engagement with the subject. Addressing these gaps in AI literacy is essential, particularly as AI’s impact on job roles within construction is expected to be transformative. The automation of labor-intensive tasks, integration of AI in project management and design, and the emergence of new career paths highlight a future where AI plays a central role in the construction workforce. Though concerns about job displacement remain, AI also presents opportunities for those with advanced technical skills in areas like AI and robotics. The evolving industry landscape underscores the need for ongoing education and training to ensure that construction professionals are equipped with both traditional expertise and AI proficiency.
Overall, students exhibit a positive outlook toward AI. A majority expressed enthusiasm, with 57% identifying as “Somewhat Optimistic” or “Very Excited”. Students’ primary interest areas include AI-powered design and modeling software, robotics for automated construction tasks, and drones for site surveying and monitoring. This generally receptive attitude among future professionals suggests a promising landscape for AI adoption in construction. However, students also voiced practical concerns about AI integration, specifically related to job security and technical competency. Addressing these concerns through targeted education will be crucial to preparing students for an AI-enabled construction environment. Furthermore, students’ focus on AI for technical and design applications rather than administrative functions indicates a practical and grounded understanding of AI’s most valuable roles in the field.