A neural network approach to assessing building façade maintability in the tropics

A model was developed to assess the maintainability of façade using neural network techniques. Inputs were derived from comprehensive studies of 570 tall buildings (more than 12 stories) through detailed field evaluation and interviews with professionals in the whole building delivery process. Sensitivity analysis showed that the most significant factors associated with façade maintainability include the system selection, detailing, accessibility and material performance.