Main Article Content
One of the key processes carried out by quality control experts to achieve the plaster clearance of the wall is the identification of mechanical, electrical, plumping (MEP) provisions. The inspection is done with the help of an approved engineering drawing. Checking provisions in a multistory build- ing is a tedious operation and is time consuming. Since detecting provisions is a repetitive activity, inspection engineer may lose interest or make mistake leading to incorrect inspection. The use of an autonomous mobile robot is demonstrated in this paper which is capable of detecting MEP provisions directly from the wall with the help of adequate algorithm. The simulation result shows the autonomous navigation and obstacle avoidance feature of mobile robot using required sensors along with navigation nodes in the customised world. Binary classification of provision is done by training Resnet-50 computational neural network model with custom made dataset.