Computer vision and sensor fusion towards better facilities management system for smart buildings

• Main Author: Arumugam, M Logaraj
• Other Authors: Li King Ho Holden
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Mechanical engineering
• Online Access: http://hdl.handle.net/10356/78602
• Institution: Nanyang Technological University

Facility management (FM) is a line of work which includes various controls to guarantee the usefulness, comfort, security, and productivity of the built sector by coordinating individuals, places, procedures, and innovation. In spite of the fact that Facilities Management (FM) is becoming undeniably important in the building environment although this industry is still in its earliest stages locally. It faces issues such as manpower needs, process efficiency and information management. Furthermore, the ever-increasing energy consumption coupled with the cost factor and the immense carbon footprint that Singapore has been facing in this modern era of demographic slowdown and economic restructuring also brings another glaring challenge to this industry. The main focus of this project is building on a previously done project under the supervision of Assistant Professor Li King Ho, Holden which was done by using sensors and combining data collected to analytics methods of Machine Learning (ML) to predict the temperature and subsequently control the air conditioner or fan. This project adds a new element of image processing by computer vision in order to recognise actual human occupancy versus data from other sensors namely a PIR sensor. It also encompasses an Internet of Things(IoT) platform which is able to remotely gather and upload data into a cloud platform instantaneously. It aims to plot a trend of how the accuracy of the Machine Learning Model can improve with the addition of this new element. It also aims to address key issues in FM such as capturing and storing information and also handling failures or maturing equipment/facilitates.