Development of an AI-Based Prototype for Integrating Design and Cost Estimation in the Early Stages of Residential Construction: A Design Thinking Approach and Adoption Validation

Abstract

Small and medium-sized construction businesses are increasingly required to provide faster, clearer, and integrated early-stage services for residential clients. However, many SME contractors still manage design briefing, preliminary cost estimation, and proposal preparation through sequential handovers, informal knowledge, and manual coordination. This article examines how artificial intelligence (AI)-based design-to-cost prototypes can support early-stage decision-making in SME residential construction. The research used an applied design research approach with Design Thinking for the development of prototypes and UTAUT3 models adapted for adoption validation. Qualitative data is collected through a focus group discussion with homeowners and the internal functions of the construction, then translated into problem formulations, How Might We question, and prototype needs. The resulting prototype integrates guided briefing, AI-based layout creation, historical cost estimates per square meter, and a design-cost compromise dashboard. The survey of 105 respondents was analyzed using PLS-SEM. The model explains a 67.8% variation in adoption intent. Social Influence and Personal Innovation significantly affect Adoption Intentions, while Performance Expectations, Business Expectations, and Enabling Conditions are insignificant. The findings show that the early adoption of AI in SME residential construction depends not only on the perception of usability, but also on organizational support and individual openness to innovation.