Testing AI Accuracy in Quantity Takeoff: A Methodological Case Study in Commercial Construction

10 pages•Published: June 2, 2026

Hooman Sadeh, Dominick Geloso and Dimitar Todorov

Abstract

This study empirically evaluated the accuracy of an AI-based quantity takeoff (QTO) platform, Togal AI, by comparing its automated measurements against contractor-produced takeoffs for a live commercial construction project. The analysis included 29 line items spanning exterior, floor, and ceiling finishes, as well as windows and doors, quantified in square feet (SF), each (EA), and linear feet (FT). Using non-parametric statistical methods—including the Wilcoxon signed-rank and Kruskal–Wallis tests—the study examined whether systematic deviations existed between AI- and contractor-generated quantities. It was found Togal AI’s overall deviations were small but statistically significant, showing a consistent underestimation for area-based quantities while achieving near-perfect alignment for count-based items. Further analysis indicated that measurement accuracy varied significantly by finish type: ceiling finishes exhibited the highest consistency with contractor data, floor finishes demonstrated moderate agreement, and exterior finishes showed the greatest deviation. The results prove that AI performs best when quantifying repetitive, clearly bounded, and orthogonal elements but becomes less reliable when interpreting irregular geometries and complex façades. The study provides the first quantitative validation of Togal AI in professional practice and concludes that while AI can accelerate takeoff workflows, estimator oversight remains essential for accuracy assurance in complex building elements.

Keyphrases: ai based quantity takeoff, computer vision, construction estimating, togal ai

In: Wesley Collins, Anthony Perrenoud and John Posillico (editors). Proceedings of Associated Schools of Construction 62nd Annual International Conference, vol 7, pages 753-762.

Links:	https://easychair.org/publications/paper/pkrRp
	https://doi.org/10.29007/g686

BibTeX entry

@inproceedings{ASC2026:Testing_AI_Accuracy_Quantity,
  author    = {Hooman Sadeh and Dominick Geloso and Dimitar Todorov},
  title     = {Testing AI Accuracy in Quantity Takeoff: A Methodological Case Study in Commercial Construction},
  booktitle = {Proceedings of Associated Schools of Construction 62nd Annual International Conference},
  editor    = {Wesley Collins and Anthony Perrenoud and John Posillico},
  series    = {EPiC Series in Built Environment},
  volume    = {7},
  publisher = {EasyChair},
  bibsource = {EasyChair, https://easychair.org},
  issn      = {2632-881X},
  url       = {/publications/paper/pkrRp},
  doi       = {10.29007/g686},
  pages     = {753-762},
  year      = {2026}}

Download PDF Open PDF in browser