Robotic solution for digitisation of indoor environment for building information modelling

Abstract

As the world welcome Building Information Modelling (BIM) as a new methodology to govern a building’s design and details, the need for digitisation of indoor spaces will increase. Prior to the digital construction of a BIM model, spatial point cloud data has to be collected. The collection of spatial point cloud data is typically done so by means of laser scanning or photogrammetry in a manual and laborious approach. In a bid to reduce manpower and save time, a robotic solution is developed and tested. This report will follow through in detail the development of the hardware and software of the solution, leveraging on past knowledge and integrating with latest cloud technology, 3D mapping frameworks and convolutional neural network. This project aims to develop a novel and intelligent approach, with the potential to benefit the civil and architectural industry.