Deployment of Digital Technology in the Management of Material Aggregate

Proper material management remains a major challenge in the Nigerian construction industry, where poor sand and aggregate measurement contributes to material waste and cost overruns. Traditional estimation methods such as visual approximation and manual tape measurement are time-consuming and error-prone, while drone, LiDAR, and photogrammetric alternatives, though more accurate, require specialized equipment and technical expertise that are often inaccessible on small and medium-scale sites in developing regions. This study aims to design and validate a low-cost, Android-compatible mobile application for real-time sand pile volume estimation using computer vision, addressing the gap left by existing tools that depend on AR or LiDAR-enabled hardware. A Flutter-based mobile interface was used for image capture, with a Python backend integrating the Segment Anything Model (SAM) for image segmentation, OpenCV for contour extraction, and geometric models for volume computation, using a calibrated reference bucket for pixel-to-metric scaling. The system was validated in controlled outdoor field trials against manually measured and calibrated-bucket control volumes across four sand pile categories. Results showed a mean estimation accuracy of 96% (mean absolute percentage error below 13% across all categories, against up to 34% for manual estimation), with an average computation time of 10 seconds. These findings indicate that the developed application offers a more accurate, consistent, and cost-effective alternative to manual estimation, and a more accessible alternative to LiDAR- or AR-dependent tools, for sand pile volume estimation in small- and medium-scale construction projects. Blockchain Technology as a Cost Management Tool in the Nigerian Construction Industry: