Compensating the lack of big data in construction industry with expert knowledge: a case study

Abstract

Due to various reasons, there is a lack of big data in the construction industry, one of the main obstacles to a broader implementation of AI. Another obstacle is adhering to analytical methods in fields more suitable for AI solutions. If appropriately used, multidisciplinary expert knowledge can compensate for these problems and enhance the application of AI techniques in construction. The case study refers to rapid earthquake loss assessment. The problem with traditional systems is their low accuracy, making them unreliable and unusable in the recovery process, which is the purpose of loss assessment systems. Low accuracy is caused by too much uncertainty in analytical and insufficient data sets to create vulnerability curves in empirical methods. The contribution of this research is designing a new kind of rapid earthquake loss assessment system using multidisciplinary expert knowledge and AI methods. The problem of small data sets was solved using the procedure of representative sampling, which makes a small sample informative and sufficient to use. The low accuracy of analytical methods is caused by assuming theoretical vulnerability relations before an earthquake. The new approach uses trained assessors to perform on-the-ground observation of actual damage on the representative sample after an earthquake. AI methods are then used to predict damage to the remaining building portfolio, which is more accurate and still rapid enough. Another contribution is using a building representation without earthquake data which eliminates the need for analytical methods, shake maps and robust ground motion sensor networks, making the proposed framework unique and applicable in any region