Prospects of Internet of Things (IoT) in Energy Sector of Bangladesh

Abstract

The intention of this study is to investigate the Prospects of Internet of Things (IoT) in Energy Sector of Bangladesh. An IoT-based Air Condition system has been considered for the study. The study has been conducted to know technical, economic and environmental prospects of IoT in energy sector in Bangladesh. Air Conditioners (ACs) have been selected for this study as it consumes higher amount of electricity than other electronic devices in both offices and residential buildings. IoT enabled ACs have been considered to know the prospects in energy sector. To achieve the stated objectives, different methods such as experiment, observation and in-depth interview have been used in this study.

Experiment is conducted to know the power consumption and impact of IoT in different temperature settings in ACs. The study has been conducted in different offices of both Dhaka and Khulna University for several months in 2019 and 2021.Experiment is carried out in summer season of Bangladesh. The study was conducted in a decentralized cooling environment where energy optimization was constrained by human comfort level. Observations are carried out to understand users preferred set temperature for ACs.

In addition to that the study endeavors to gain insights into the prospects of Internet of things in Energy sector of Bangladesh through qualitative in-depth interview. In-depth interview has been administered among the experts from power development board, Air conditioner manufacturer and University. They were from west zone power distribution, Dhaka and Khulna University and Walton Bangladesh. The study uses auto-regressive machine learning and other techniques to capture and record data from various sources. The smart system predicts which temperature setting can meet the desired energy saving given the present temperature and humidity of indoor and outdoor, specific day and time of the week. An auto-regressive machine learning model was trained to predict optimal action for the given level of cost saving. Collected data have been carefully organized and analyzed using MS excel and Python programming language.

The result shows, LSTM model could predict the discrete optimal control signal for 3 air- conditioner with 94% accuracy. Although the work was not applied to centralized air- conditioning system, the author believes, with extensive amount of data, the model can learn optimal saving strategies in centralized setting.

The study reveals that IoT enabled ACs save energy, reduce 𝑪𝑂2 emission significantly without compromising users’ comfort level. Major findings of the of the study are: significant level of savings such as 26.1%-30% energy consumption reduction and nearly 129885 tons of 𝑪𝑂2 emission reduction is possible with the help of IoT enabled ACs in energy sector in Bangladesh. The study is significant as the prediction model showed enormous potential to save energy as well as reduce the carbon emission in a significant way. It also reveals that savings of energy through use of IoT in ACs varies as per the behavior of the users. At the end of this thesis paper policy recommendations, future research direction and limitations of the study are articulated.