Real-time Asset Information Modelling (rtAIM) Protocol for Highways

To operate successfully, an asset-centric organisation must recognise the need of visually validating the placement of its constructed assets. This is important to organisations but there are limitations and challenges. Building Information Modelling (BIM) has been widely adapted, and its establish methods and technologies offers some potentials in benefiting the highways sector. Hence this study involved the development of a BIM-based protocol and decision framework for real-time collection, validation, and handover of attribute data for National Highways' major schemes, using machine learning from images collected from drone flights. The study has integrated qualitative and quantitative approaches at rigorous and extensive stages, following the principles of critical realism theory and exploratory sequential mixed methods. The qualitative research further informed the design of a questionnaire which was used to elucidate broader industry experts’ perspective that ultimately guided the design, development and validation of machine learning-enabled real time asset information modelling. A new protocol to overcome these limitations by applying Machine Learning algorithms for Mask RCNN (Region based Convolutional Neural Networks) and recognising the assets of roads with geospatial images obtained from drones. The prototype looked at a linear asset, as they are the most difficult to capture. Using 150 images of the chosen asset type, these were labelled then processed using machine learning which then highlighted the assets it had learnt, allowing the output to be sent to client databases in there required file format. As proposed in the framework and validated through a case study the prototype effectively showcased how drone photogrammetry, powered by Machine Learning, a subset of Artificial Intelligence and Building Information Modelling (BIM) processes, can capture assets in real-time. This process reduces the time for redlining and negates the need for on-site surveys, adding value and reducing programme time.