Implementasi BIM Berbasis Artificial Intelligence untuk Optimasi Perencanaan Struktur Atas Gedung Katering Pondok Pesantren Kedunglo Al Munadhoroh

Abstract

ABSTRACT

Conventional structural design based on two-dimensional drawings and manual calculation is prone to data inconsistency, quantity take-off errors, and long processing time. This study aims to implement Artificial Intelligence (AI)-based Building Information Modelling (BIM) for optimizing the superstructure design of the Kedunglo Al Munadhoroh Islamic Boarding School Catering Building and to compare the cost budget plan (RAB) between the initial design and the optimization results. This applied research with a quantitative-comparative approach integrated Autodesk Revit as the modelling platform, six Dynamo scripts as a rule-based automation system, Autodesk Robot Structural Analysis Professional for structural analysis, and AI as a decision support system. The five-story, 22 m high building was designed as a Special Moment Resisting Frame, Risk Category III, and Seismic Design Category D, referring to SNI 2847:2019, SNI 1726:2019, and SNI 1727:2020. The AI generated ten dimensional variations filtered into three alternatives, namely D1 (Existing), D2 (Minimum), and D3 (Maximum), which were then evaluated through a four-criteria weighted decision matrix. The results established D1 as the most optimal alternative with a weighted score of 3.60 (72%), outperforming D2 (3.50; 70%) and D3 (3.40; 68%). The RAB of D1, IDR 4,433,656,450.00, was 6.2% above D2 yet 16.3% more economical than D3; the optimization therefore validated the initial design as the best balance between structural safety and cost efficiency.