AI PERCEPTION IN CONSTRUCTION
Human-Centered AI Perception Prediction in Construction: A Regularized Machine Learning Approach for Industry 5.0
Industry 5.0 emphasizes human-centered AI integration, but successful adoption hinges on workforce perception. This research validates a machine learning framework for predicting AI perceptions and expected impacts in construction, achieving meaningful accuracy with limited data. Our approach enables targeted interventions for effective digital transformation.
Key Predictive Performance Metrics
Our framework delivers robust performance in predicting AI perception, even with the data constraints typical of specialized industrial surveys. The metrics below demonstrate effective generalization and statistical validity for human-centered AI adoption in construction.
0.501
Cost Reduction Perception (Regression)
0.681
Weighted Average (Classification)
51
Valid Survey Responses (Data Points)
1.26
Company Size Perception Gap
Deep Analysis & Enterprise Applications
Select a topic to dive deeper, then explore the specific findings from the research, rebuilt as interactive, enterprise-focused modules.
This research introduces a machine learning framework to predict AI-related perceptions in construction, overcoming small sample constraints. Our approach focuses on data engineering, careful model selection, and rigorous validation to ensure robust and generalizable insights. This systematic methodology allows for meaningful predictions without relying on large datasets or complex architectures, making it suitable for specialized industrial contexts.
Enterprise Process Flow: AI Perception Prediction Framework
Phase 1: Data Engineering (Target Simplification, Feature Reduction)
→
Phase 2: Model Selection (Low-variance, Regularized Models)
→
Phase 3: Validation Strategy (Leave-one-out Cross-Validation)
→
Phase 4: Ordinal Target Handling (Class Weighting)
Our analysis identifies key predictors of AI perception and validates the superiority of regularized models in small-sample contexts. The Lasso regression model achieved an R² of 0.501 for cost reduction perception, demonstrating its effectiveness. Classification tasks, addressing various AI impact dimensions, yielded a weighted F1-score of 0.681. These results underscore that simplicity and regularization are crucial for reliable predictions with limited data, preventing overfitting that often plagues more complex models.
| Model | MAE | RMSE | R² | Key Characteristics |
|---|---|---|---|---|
| Lasso (L1 regularization) | 0.551 | 0.709 | 0.501 |
|
| Ridge (L2 regularization) | 0.559 | 0.714 | 0.494 |
|
| Elastic Net (L1 + L2) | 0.556 | 0.715 | 0.493 |
|
| Random Forest (depth = 3) | 0.568 | 0.770 | 0.412 |
|
| Gradient Boosting (depth = 3) | 0.666 | 0.845 | 0.292 |
|
Key Predictive Driver
Our research identified the single most influential factor shaping AI perception in the construction industry.
Personal AI Usage Strongest Perception Predictor (Beta: 0.359), aligned with "perceived usefulness."Organizational scale significantly influences AI adoption perceptions. Larger companies generally demonstrate more favorable attitudes due to greater resource availability, technology infrastructure, and formal training opportunities. This insight helps organizations tailor their AI strategies to specific company sizes, ensuring efforts are effectively matched to internal capabilities and employee readiness.
Company Size & AI Perception: A Key Determinant
Our analysis reveals a significant 1.26-point perception gap between smaller (micro & small) and larger (medium & large) construction companies. Larger organizations consistently show more positive AI perception scores (mean 4.12-4.31) compared to smaller ones (mean 2.75-3.17), with a statistically significant effect (p < 0.001, Cohen's d = 1.47).
This substantial difference highlights that organizational resources, existing technology infrastructure, and formal training opportunities are critical drivers for AI adoption attitudes. For smaller companies, this means AI solutions should prioritize ease of implementation, minimal infrastructure requirements, and rapid time-to-value to overcome perception barriers effectively.
The predictive framework offers concrete guidance for effective AI adoption. We emphasize hands-on AI exposure over traditional training, tailored communication strategies for different workforce segments, and the importance of foundational digital competencies. By prioritizing interventions based on predicted impact, organizations can efficiently allocate resources and mitigate resistance, ensuring a human-centered transition to Industry 5.0.
Enhanced Estimation Stability
Our framework significantly improves the reliability of parameter estimation by optimizing the balance between sample size and feature count.
n/p ≈ 7.3 Optimized Samples-Per-Feature Ratio for Stable Parameter Estimation.While our framework demonstrates strong performance for small samples, inherent limitations include sample size (n=51), geographical specificity, and reliance on self-reported data. Future work will focus on expanding the dataset to 150-200 respondents to further enhance prediction accuracy (aiming for R² ≈ 0.65-0.70) and integrating psychological and contextual factors. Longitudinal and multi-source data collection will refine our understanding of perception dynamics and intervention effectiveness.
Achievable Prediction Ceiling
With expanded datasets and further refinements, the predictive power of AI perception models can reach new heights.
R² ≈ 0.65-0.70 Realistic R² Ceiling with Substantially Expanded Datasets.Advanced ROI Calculator for AI Adoption
Estimate the potential cost savings and reclaimed work hours from optimizing AI perception and adoption within your organization. Adjust the parameters to see a customized impact forecast.
Estimated Annual Cost Savings
$40,000
Estimated Annual Hours Reclaimed
1,000
Your AI Adoption Roadmap
Based on the predictive insights, here’s a strategic roadmap for human-centered AI integration, focusing on maximizing perception and impact at each stage.
Assess Digital Readiness
Evaluate current ICT utilization, digital competencies, and existing AI exposure within your workforce. Identify key segments and roles for targeted interventions to build a strong foundation for AI acceptance.
Pilot Targeted AI Solutions
Implement structured AI tool trials on pilot projects, focusing on hands-on experience over theoretical training. Prioritize solutions with clear, rapid time-to-value, especially for smaller companies and specific operational tasks.
Customize Training & Communication
Develop role-specific communication strategies addressing concerns like job displacement and skill obsolescence. Design training programs that emphasize human-AI collaboration and practical application, ensuring relevance and perceived usefulness.
Integrate Workflows & Scale
Ensure foundational digital workflows (e.g., digital document management, collaborative scheduling) are robust before enterprise-wide AI deployment. Systematically integrate AI tools into daily operations, demonstrating tangible benefits across functional areas.
Monitor, Evaluate & Refine
Continuously track AI perception dynamics and intervention effectiveness. Use model coefficients to prioritize resource allocation, adapting strategies based on real-world feedback and evolving organizational needs to sustain positive adoption patterns.
Ready to Optimize Your AI Strategy?
Our human-centered AI prediction framework provides actionable insights tailored to your organization. Let's discuss how to leverage these findings for a successful, perception-aware AI implementation in your construction enterprise.
Ready to Get Started?
Book Your Free Consultation.
Let's Discuss Your AI Strategy!
Lets Discuss Your Needs
Select a Date
Sun
Mon
Tue
Wed
Thu
Fri
Sat