The state of the testing software designated for information models of construction projects and prospects for their application

Abstract

Introduction. The problem of automated testing in the process of designing construction facilities has been solved at the international level for more than 40 years. Earlier articles had overviews of design verification systems, presented in the form of information models (IM) of construction projects. However, over the last few years the process of digitalization of the construction industry has become more intense, and new countries, including Russia, have been more actively involved in it. Hence, new methodological approaches to individual stages of verification, programmes and systems, not described in earlier reviews, have appeared. At the same time, many previously developed systems have been modified or, conversely, have ceased to exist. The purpose of this article is to evaluate the current state of IM verification systems for construction projects, taking into account the changes that have taken place over the last few years, and to determine the prospects for their further development. Materials and methods. To determine the current state of systems for testing the IM of construction facilities, the co-authors selected and analyzed the foreign and Russian literature and information sources in the field of testing the IM of construction facilities. The results of earlier reviews were also taken as the benchmark. Results. The co-authors made a list of currently used replicable commercial solutions for information model testing, which are classified according to their designation and a per-country list of information model testing systems, with the status identified for each system. The co-authors identified the development areas in respect of verifying international models of construction projects of international scale. Development areas in the field of verification of informational models of construction facilities at the Russian Federation level were also outlined. Conclusions. Presently, there is still a problem of converting regulatory requirements into the machine-readable format to ensure their compliance with Russian and international standards. Therefore, the main direction for the further development is the study the potential of artificial intelligence in the processing of regulatory requirements written in a natural language. Nevertheless, the application of neural networks requires the availability of data for training, which suggests the need for a certain amount of manually marked regulatory documents in advance.