Structural Health Monitoring aided by neural networks

Despite of the age of the building and the nature of materials used in it, it is important to retain the reliability and service life of structures. This is established by proper monitoring of various structural members. A distress may occur in a structure due to numerous reasons from shrinkage cracks due to hot weather at the time of concreting to chloride induced corrosion or creep deflections at later stages. Due to the diversity in the factors that cause structural distress, proper recognition becomes quintessential. The R&D in the field of structural design is progressing day by day. So is the amendments in the codal provisions related to it. This paper intends to contemplate the available latest structural health monitoring mechanisms based on pertinent literature. The main emphasis of the paper is on the need for proper monitoring of existing and upcoming structures and also to employ proper monitoring techniques. The limitations in existing knowledge and lack of codal provisions can be tackled by suitable monitoring techniques. For forthcoming constructions, monitoring mechanisms could be used for validating design assumptions and software analysis results. These could be achieved by proper incorporation of building information modelling (BIM), sensors and neural networks. There have been serious structural failures, which could have been avoided by providing necessary retrofitting based on inspection. This flaw of not inspecting and monitoring of constructed structures could be tackled by structural health monitoring techniques and this could be made convenient by use of neural networks. Thus, making structural maintenance economical, as remedy is started at initial stage itself.