Plag
Plag
Plag
Page 1 of 3
project. Finally, our sincere thanks to our parents, family members and friends
for their continuous support, inspiration, and encouragement without which
this project would not have been a success. Vinit Kumar Khevito Shohe
Dipanshu Abstract The Manhole Gas Detection System using ESP32 and C++
is an outstanding solution for enhancing operational safety and efficiency
within urban infrastructure by continuously monitoring hazardous gases'
level in manholes. Manholes allow access to a variety of critical underground
utilities that may be subject to dangerous accumulation of gases like
methane and hydrogen sulfide. One should install a reliable and accurate
real-time gas detection system, which can provide early detection for
accident avoidance and save the lives of people in time. It is powered by the
ESP32 microcontroller, offering dual-core processing power and integrated
with Wi-Fi and Bluetooth connectivity. Coupling a high-performance gas
sensor together with the ESP32 enables the continuous monitoring and
analysis of gas concentrations, providing real-time data to a monitoring
station situated at some central facility. This system efficiently does all the
processing and communication needed, alerting in real-time to hazardous
conditions and providing an effective response through C++ programming.
The key features of the system include scalability to work with multiple
manholes and in a variety of urban settings while offering cost-effectiveness
by providing a high-performance solution at a much lower cost compared to
gas detection systems traditionally deployed. The system will be designed
according to safety standards and regulatory requirements so that both
compliance and safety protocols are improved. It shows its reliability and
accuracy in a variety of environments, thus highly useful in managing urban
infrastructure, industrial facilities, and emergency scenarios. The ESP32-based
Manhole Gas Detection System is the first milestone and proves leading
technology for gas detection that enhances safety, efficiency, and
management of underground infrastructure. List of Figures / charts Sl No.
Description Page Number Figure - 3.1 Methodology Diagram 15 Figure - 3.2
Power Bi Dashboard of the dataset 18 Figure - 3.3 Comparison bar graph of all
the machine learning algorithms 19 Figure - 3.4 Analysing the target value. 21
Figure - 3.5 Analysing the sexfeature. 21 Figure - 3.6 Analysing the 'Chest Pain
Type' feature. 23 Figure - 3.7 Analysing the resting electrocardiographic results
feature 25 Figure - 3.8 Analysing the ' exercise induced angina' feature. 26
Figure - 3.9 Analysing the slope of the peak exercise ST segment feature. 27
Figure – 3.10 Visual representation of user interface. 35 List of Schemes/
Algorithms Sl No. Algorithm Description 1 Pandas Python library for data
manipulation and analysis, providing powerful data structures. 2 NumPy
Fundamental package for scientific computing in Python, facilitating array
operations and mathematical functions. 3 Scikit-learn Python machine
learning library with simple and efficient tools for data mining and analysis.
Many classification algorithms were used and evaluated. Like Logistic
regression, Naïve bayes, KNN, Decision tree, Random Forest. 4 Stream-Lit
Python framework for creating interactive web applications for machine
learning and data science projects. 5 Matplotlib Python 2D plotting library for
creating static, interactive, and animated visualizations in various formats.
Page 2 of 3
CHAPTER 1: INTRODUCTION Safety and reliability of underground systems in
contemporary urban infrastructure are very important to the well-being of
both the general public and maintenance personnel. Typical manholes,
usually being entry points to most of the underground utilities, may
accumulate hazardous gases. These include methane and hydrogen sulfide,
associated with significant risks of explosion, health hazards, and operational
disruptions. These risks call for effective and reliable detection systems that
enable early warnings and timely interventions.
Sources
5% Plagiarized
Submitted in Partial Fulfilment of The Requirements For The Award of Degree
of. Uploaded by. Jyoti Arvind Pathak. 100%(1)100% found this document useful
(1 ...
https://www.scribd.com/document/478375157/Final-report
3% Plagiarized
by V Kathiresan · 2022 · Cited by 6 — The classification algorithms of machine
learning like logistic regression, naïve Bayes, KNN, Decision Tree, Random
Forest, and SVM are suitable for ...
https://ieeexplore.ieee.org/document/9740744
Page 3 of 3