INTRODUCTION
OVERVIEW OF THE SMART WAREHOUSE SYSTEM
Recent advancements in Industry 4.0 and the Internet of Things (IoT) have sparked considerable interest in Vietnam's IoT System market As technological innovation accelerates, IoT System adoption is on the rise in many developed nations Notably, Vietnam's two largest cities rank among the top 10 fastest-growing urban areas for economic and real estate development, according to Jones Lang Lasalle's City Momentum Index 2020 Although still emerging, Vietnam's IoT System market presents significant investment opportunities, driven by rapid urbanization and the rise of mega-cities.
Figure 1.1 The projected revenue for the IoT System market in Vietnam is expected to be high
Factors influencing the potential development of the IoT System market
Vietnam's IoT System market is rapidly expanding due to urbanization, widespread smartphone adoption, and a significant millennial demographic The country boasts the highest global growth rate in smartphone usage, with tech-savvy millennials driving demand for innovative solutions As reported by Vnexpress, many users appreciate the convenience, safety, and energy efficiency of IoT Systems However, concerns arise regarding the usability of these technologies for older family members, as multi-generational households are common in Vietnam, posing challenges for the elderly in adapting to smart devices.
According to an Acis representative on EinPresswire, three primary barriers hinder individuals from adopting IoT System technology First, a traditional mindset persists, as users are accustomed to using contactors for controlling home electrical devices, creating a psychological barrier to transitioning to smart alternatives Second, the cost of smart technology remains a significant obstacle; while domestic brands offer basic IoT packages ranging from $1,500 to $4,500, a fully functional IoT System with advanced features can exceed $25,000, which is a substantial investment given the country's average GDP per capita of $2,715 and a monthly income of approximately $180 Lastly, the complexity of installation adds to the reluctance of potential users to embrace IoT solutions.
Real estate developers in Vietnam emphasize the necessity of constructing a significant number of homes utilizing modern building techniques to facilitate the integration of IoT system technology Furthermore, for homes to achieve true smart status, there must be an enhanced and more reliable network connectivity throughout the country.
IoT System market trends are expected in the coming years
In urban centers like Hanoi, Ho Chi Minh City, and Da Nang, the adoption of IoT systems is rapidly gaining traction among modern residents, particularly millennials who are adept with the latest technology This burgeoning market is also appealing to young families seeking to leverage technology for an enhanced quality of life.
The growing popularity of IoT Systems is prompting developers to integrate smart technology into construction projects, despite the challenges posed by high costs and increasing demand As part of the Industry 4.0 trend, the necessity for IoT Systems and smart buildings is becoming more pronounced Beyond technological advancements, these systems must also prioritize environmental sustainability, energy efficiency, and safety standards To ensure long-term sustainable development, it is crucial for developers in Vietnam to create genuinely smart apartments, moving beyond mere marketing jargon.
The Internet of Things (IoT) is revolutionizing household electronics, enhancing everything from entertainment systems to smart home management Devices like smart TVs and washing machines are rapidly integrating IoT technology, leading to a more connected home environment This surge in IoT adoption is primarily fueled by the rising demand for specialized microcontroller units and the innovation of versatile System on Chips (SoCs) tailored for various applications.
In July 2021, LifeLong ventured into the Internet-of-Things (IoT) industry, introducing solutions across three key categories: sensors and detectors, lights and fans, and utilities and devices These innovative offerings aim to tackle common challenges in managing electrical devices, positioning LifeLong for substantial market growth in the coming years.
The emergence of smart cities presents significant opportunities for the Internet of Things (IoT) in energy management, waste reduction, and infrastructure development Smart homes, an integral aspect of this concept, offer numerous advantages Additionally, advancements in 5G technology will enhance IoT capabilities by enabling a high volume of simultaneous connections, facilitating the growth of networks in densely populated urban areas characterized by high mobility.
Figure 2.2 Sales of Smart Home Device, in billion USD, United States, 2016-2020
Smart home devices and voice-activated assistants, such as Amazon Echo and Google Home, have revolutionized convenience in households globally In 2020, Amazon Web Services (AWS) reported over 500 million connected devices, serving more than 10,000 customers through their IoT infrastructure.
Smart devices are increasingly popular, featuring advanced connectivity solutions For instance, LG's smart refrigerators enable users to pre-order ice while purchasing ice-dependent items and assist in checking product expiration dates, helping consumers optimize their shopping lists.
Figure 3.3 Internet of Things (IoT) Devices Market - Growth Rate by Region (2021 - 2026)
In summary, the implementation of IoT systems is crucial as it significantly improves user experience, adds value to everyday life, enhances living standards, and saves time, money, and energy As awareness of these benefits grows, individuals are likely to embrace IoT solutions, overcoming any hesitations about integrating new technology into their homes.
PROBLEMS
In today's fast-paced technological landscape, "IoT Systems" have emerged as a significant global trend, transforming traditional living spaces into optimized and convenient environments These systems enable devices and systems to connect and interact seamlessly through the Internet of Things (IoT), enhancing the overall living experience.
The swift advancement of technology, coupled with enhanced utility and energy efficiency, has significantly boosted the global adoption of IoT systems These systems offer numerous advantages to residents, including the ability to control lighting, temperature, and security, as well as facilitating smart management and monitoring of energy consumption.
Implementing IoT systems presents several challenges, including concerns about the safety and security of personal data Additionally, it's essential to evaluate whether the energy savings provided by these systems justify the initial investment costs Furthermore, there is a risk that smart living environments may lead to isolation and reduced community interaction.
Figure 1.4 Smart Warehouse with Internet-of-Things Technology
This report examines the development, potential, and limitations of IoT system technology by analyzing existing systems and real-life applications It evaluates the impact of IoT on modern life and offers solutions and recommendations to enhance the benefits of IoT technology, aiming to create a convenient, safe, and sustainable living environment.
As technology evolves, IoT systems are emerging as a crucial component of our future rather than just a passing trend By exploring and comprehending IoT technology, we can unlock its potential to create a smarter and more advanced living environment for the future.
IoT System have many advantages over traditional homes, but there are also some drawbacks to consider Below are some pros and cons of IoT systems compared to traditional homes:
IoT systems offer homeowners the convenience of remote management through mobile apps and web interfaces, enabling effortless control of lights, temperature, security systems, and various other devices This seamless integration enhances the ease of use, making home automation simpler than ever before.
Utilizing IoT technology enables automatic adjustments of temperature, lighting, and various devices through data collected from sensors This smart management not only conserves energy but also leads to significant reductions in electricity bills.
IoT systems enhance security and monitoring for homeowners, allowing them to oversee and manage security systems, surveillance cameras, and sensors online This connectivity ensures the safety of both family members and property.
IoT systems offer exceptional integration and expandability, allowing homeowners to seamlessly connect a variety of devices and services By linking household appliances, sound systems, lighting, and electronics, individuals can create a personalized and optimized living environment tailored to their unique needs.
Implementing an IoT system involves a significant upfront investment, often exceeding that of traditional homes This initial cost encompasses the purchase and installation of devices, network infrastructure, software, and ongoing maintenance and upgrades.
IoT systems present significant challenges regarding security and privacy, as the connection of devices and personal data to the internet can lead to vulnerabilities This connectivity heightens the risk of network security breaches and intrusions, making it essential to address these concerns effectively.
To maximize the benefits of an IoT system in a home, a constant internet connection is essential Any disruptions in the internet service can hinder the ability to control and manage connected devices effectively.
Potential for malfunctions and instability: IoT systems using IoT may encounter technical issues, software bugs, or instability This can cause inconveniences and require regular checks and repairs
When considering the implementation of an IoT system, it is crucial to weigh its advantages and disadvantages to maximize its benefits while ensuring the safety and sustainability of your home and family.
RESEARCH PURPOSE
The purpose of this research is to design and implement an indoor monitoring, control and warning system using IoT technology, with the hardware component being ESP32
The system incorporates facial recognition technology for secure door unlocking and device control, while utilizing the RF21L01 communication protocol for effective monitoring of sensor data Additionally, it connects to Firebase for seamless integration with both the website and mobile app The website features advanced data visualization capabilities, enabling users to analyze sales trends and access weather forecasting information.
The research objectives are as follows:
- To design and implement an indoor monitoring, control and warning system using IoT technology with an ESP32 as the hardware component
- To implement facial recognition for door unlocking and device control
- To integrate sensor data using the RF21L01 communication protocol
- To connect the system to Firebase for integration with the Website and App
- To develop a website with features for data visualization of sales and weather forecasting
RESEARCH METHODS
The indoor monitoring, control and warning system using IoT technology with ESP32 as the hardware component will be designed and implemented using the following research methods:
- Conduct a thorough literature review of existing IoT technologies and facial recognition systems to identify the most appropriate hardware and software components for the system
- Design and develop the hardware components of the system, including the ESP32 microcontroller, RF21L01 communication protocol, sensor inputs, and motor control
- Implement facial recognition software using machine learning algorithms to enable door unlocking and device control
- Develop a data acquisition system to collect data from sensors and send it to Firebase for storage and analysis
- Design and develop a user-friendly web interface that allows users to monitor the system's status, control devices and access data from anywhere
- Implement data visualization techniques to provide real-time data analytics and forecasting of sales and weather
- Test the system's performance under various conditions to validate its reliability and accuracy
- Analyze the results of the tests and fine-tune the system's design and operation to optimize performance
- Document the system's design, implementation, and testing procedures for future reference
- Disseminate the findings of the research through publications, presentations, and other communication channels to promote the benefits of IoT technology for indoor monitoring, control and warning systems.
LITERATURE REVIEW
I2C
I2C (Inter-Integrated Circuit) is a synchronous serial communication protocol developed by Philips Semiconductors It is used to transmit and receive data between integrated circuits using only two signal transmission lines
Data bits are sent one by one at regular intervals established by a clock signal
The I2C bus is commonly used for peripheral communication for various types of ICs, such as microcontrollers, sensors, EEPROMs, and more
I2C uses two signal transmission lines:
SCL (Serial Clock Line): generates clock pulses sent by the Master
SDA (Serial Data Line): transmits and receives data I2C communication involves data transmission between Master and Slave devices
The Master device is typically a microcontroller that controls the SCL signal and sends/receives data or commands through the SDA line to other devices
The devices receiving the data and commands from the Master are called Slave devices These devices are typically ICs or even microcontrollers
In an I2C bus configuration, the Master and Slave devices are interconnected, with both the SCL and SDA lines functioning in Open Drain mode This setup allows any device on the bus to only pull these lines down to a LOW level, while a pull-up resistor (ranging from 1 to 4.7 kΩ) is required to maintain a HIGH level by default and prevent bus conflicts where one device might attempt to pull the line up while another pulls it down.
Standard mode with a speed of 100 kBit/s
Low-speed mode with a speed of 10 kBit/s
Unlike SPI communication, which supports only a single Master device, I2C communication enables data exchange between multiple Master devices and a Slave device This capability can complicate the process, as the Slave device may receive simultaneous data frames from various Masters, potentially resulting in conflicts or errors in the data received.
To avoid this, when working in this mode, each Master device needs to detect the state of the SDA line
When the SDA line is at a LOW level, it indicates that another Master device is in control of the bus, requiring the current Master device to wait until the transmission is finished.
Conversely, if the SDA line is at a HIGH level, it means that the SDA line is safe to use and is available for data transmission.
SPI
SPI (Serial Peripheral Interface) is a serial peripheral interface developed by Motorola
It is a synchronous standard for full-duplex data transmission, meaning that data can be transmitted and received simultaneously at any given time
As a synchronous communication protocol, all processes are synchronized with the clock signal generated by the Master device, so there is no need to worry about data transmission speed
SPI (Serial Peripheral Interface) is widely utilized for communication with various devices, including EEPROM memory, Real-Time Clocks (RTC), audio integrated circuits (ICs), temperature and pressure sensors, as well as memory cards like MMC and SD cards, and even other microcontrollers.
SPI uses four communication lines, sometimes referred to as the "4-wire" communication standard These four lines are:
- SCK (Serial Clock): The Master device generates the SCK signal and provides it to the
The slave device plays a crucial role in maintaining synchronization during SPI communication Each pulse on the SCK line transfers a single bit of data, which minimizes errors and facilitates high-speed transmission.
MISO, or Master Input Slave Output, refers to a communication line where the Slave device generates a signal that is received by the Master device For proper functionality, it is essential to connect the MISO line between these two devices.
MOSI, which stands for Master Output Slave Input, is a crucial signal in communication between devices It is generated by the Master device and received by the Slave device, necessitating a direct connection of the MOSI line between them for effective data transfer.
The Slave Select (SS) pin is crucial for communication in a Master-Slave configuration, as it allows the Master device to choose a specific Slave device by pulling the corresponding SS line to a LOW level (0) This pin is also known as Chip Select (CS), and users can configure any GPIO pin on the microcontroller as the SS pin by setting it to Output mode.
Each Master or Slave chip has an 8-bit data register
The transmission process between the Master and Slave occurs simultaneously after 8 clock cycles, and one byte of data is transmitted in both directions
The data exchange process begins when the Master generates a clock signal from the Clock Generator and pulls the Slave's SS line down to a LOW level to initiate data transmission
During each clock pulse, the Master transmits 1 bit from its Shift Register to the Slave's Shift Register through the MOSI line, while simultaneously receiving 1 bit from the Slave via the MISO line Consequently, after 8 clock cycles, the complete transmission and reception of 1 byte of data is achieved.
Since the data of the two registers are exchanged with each other, the data exchange speed is fast and efficient.
RADIO (RF) COMMUNICATION PROTOCOL
Advantages of Radio Frequency (RF) Waves:
RF waves are utilized in numerous industries due to their versatility, playing crucial roles in wireless communication, broadcasting, radar systems, remote control devices, and medical imaging This broad range of applications highlights the significance of RF waves in various scenarios.
Long-distance communication is effectively facilitated by RF waves, which can travel great distances with minimal signal loss This characteristic is especially beneficial in situations where wired connections are either impractical or expensive, such as in remote locations or mobile communication scenarios.
RF waves exhibit excellent penetration capabilities, enabling them to traverse obstacles such as walls, buildings, and foliage This attribute makes RF waves particularly suitable for indoor wireless communication, where overcoming physical barriers is essential for maintaining signal strength and connectivity.
RF-based communication systems offer significant scalability by supporting multiple devices at once, making them ideal for large-scale deployments This capability is particularly important in high-density environments, such as bustling cities or populated regions, where numerous users require reliable connectivity.
Disadvantages of Radio Frequency (RF) Waves:
RF waves function within a restricted frequency spectrum, leading to limited bandwidth availability As the demand for wireless connectivity increases with more devices and applications, this available bandwidth can become congested, causing slower data transmission speeds.
RF waves are highly susceptible to interference from nearby RF devices or those operating on overlapping frequencies, which can significantly degrade signal quality and communication reliability To effectively mitigate interference, careful frequency planning and management are essential.
Security vulnerabilities in RF communications arise when unauthorized parties intercept RF waves, threatening the privacy and integrity of transmitted data To safeguard these communications, implementing robust encryption and authentication measures is essential to prevent unauthorized access and enhance overall security.
RF waves can partially penetrate obstacles, but maintaining a clear line of sight is essential for optimal performance Significant barriers and terrain can impede signal propagation, leading to decreased range and quality in RF communication.
FACE DETECTION
Face detection is a computer vision technology that identifies and locates human faces in images or videos Utilizing advanced algorithms, it analyzes visual data to recognize key facial features, including the eyes, nose, mouth, and chin.
Face detection is often used as a preliminary step in various
Face detection is a multi-step process that includes image acquisition, preprocessing, feature extraction, and classification This process primarily relies on machine learning algorithms, utilizing techniques such as artificial neural networks, support vector machines, and decision trees to accurately identify faces in images.
Face detection is a powerful technology with numerous practical applications Unlike the mature fields of fingerprint and iris recognition, face recognition continues to face challenges, making it a compelling area for research It offers a richer data source, as human faces can be found in various images and videos online, and requires less controlled interaction compared to fingerprint or iris recognition, which necessitates cooperation in a controlled environment.
Figure 2.6 The Face Detection Algorithm Set to Revolutionize Image Search
Face recognition methods are categorized based on various criteria, with still image-based face recognition (FR) using 2D input data being the most prevalent However, the future of face recognition may shift towards 3D FR, which utilizes multiple 2D cameras to generate 3D data, offering more accurate and reliable results This 3D FR can be further divided into two approaches: processing image data and analyzing video data.
Figure 2.7 Exampel Facial Recognition APIs & Software
Face recognition methods can be categorized into three main types: global approaches like Eigenfaces-PCA and Fisherfaces-LDA, local feature-based methods such as LBP and Gabor wavelets, and hybrid approaches that integrate both global and local features Among these, local feature-based methods have demonstrated superior performance in uncontrolled conditions The evolution of face recognition technology can be viewed as a continuous journey focused on the advancement of feature extraction techniques used in feature-based systems.
Face detection involves identifying human faces in images or videos, a task complicated by variations in facial features like size, orientation, and lighting Despite these challenges, advancements in computer vision and machine learning have led to the creation of precise and dependable face detection algorithms.
Face detection involves multiple stages, starting with image acquisition, where a camera captures an image or video frame The captured image undergoes pre-processing to improve quality and minimize noise, which includes resizing, filtering, and adjusting brightness and contrast.
The feature extraction process involves analyzing images to detect key facial features, including the eyes, nose, mouth, and chin Various techniques are utilized for this purpose, such as Haar cascades, Local Binary Patterns (LBP), and advanced deep learning algorithms like Convolutional Neural Networks (CNNs).
Over the years, face detection methods have significantly advanced, driven by improvements in computer vision and machine learning technologies This article explores various widely-used face detection techniques, highlighting their respective strengths and weaknesses.
Figure 2.8 Different Face Recognition Techniques
Haar cascades, introduced by Viola and Jones in 2001, are a widely used method for face detection that utilizes simple rectangular features to identify key facial elements like the eyes, nose, and mouth By combining these features, a classifier is created to efficiently detect faces in images, enabling real-time performance Despite their computational efficiency, Haar cascades face challenges in recognizing faces with non-frontal poses and in low-light environments.
LBP, or Local Binary Patterns, is a widely used face detection technique that identifies facial features by analyzing the texture of images It operates by comparing the intensity values of adjacent pixels and generating a histogram from these values This method is effective in handling lighting variations and can detect faces in non-frontal positions However, LBP may experience high false positive rates and can struggle with low-resolution images.
Convolutional Neural Networks (CNNs) are a powerful deep learning approach for face detection, demonstrating remarkable accuracy improvements By training on extensive datasets of facial images, these algorithms effectively learn to identify facial features, enabling them to detect faces in new images across various poses and lighting conditions Despite their high accuracy, CNNs demand considerable computational resources, which can limit their applicability for real-time face detection scenarios.
The Viola-Jones algorithm is a popular method for face detection that integrates Haar cascades with machine learning It employs a series of weak classifiers to identify faces within images, offering computational efficiency and real-time detection capabilities Despite its advantages, the algorithm can experience high false positive rates and may struggle with low-resolution images.
• Histogram of Oriented Gradients (HOG):
HOG (Histogram of Oriented Gradients) is a widely used face detection technique that identifies facial features by analyzing the gradient orientation of an image This method involves segmenting the image into small cells and calculating the gradient orientation for each, which is then compiled into a histogram HOG is effective in varying lighting conditions and can detect faces across diverse poses However, it is prone to high false positive rates and may struggle with occluded faces.
PCA is used to reduce the dimensionality of data by transforming the original variables into a new space with distinct components It reduces complexity by removing extra dimensions
LDA searches for a new dimensional space in which the separation between layers becomes maximum It is often used after PCA to optimize discrimination between different faces
Eigenface employs Principal Component Analysis (PCA) to identify the key features, known as eigenvectors, within a collection of faces, which are then represented as "eigenfaces." New facial images are expressed as a linear combination of these eigenfaces, allowing for efficient representation Additionally, eigenvalues are utilized to minimize data size and enhance processing speed, making the system more effective for facial recognition tasks.
FIREBASE
Firebase is a powerful platform for mobile and web application development, offering a comprehensive set of tools and services that enable developers to create high-quality applications efficiently Initially launched by Firebase Inc in 2011, it was acquired by Google in 2014, further enhancing its capabilities and reach in the developer community.
Firebase provides a comprehensive suite of features and services such as a real-time database, authentication, cloud storage, hosting, and cloud functions, enabling developers to create scalable, reliable, and secure applications with ease.
Firebase's real-time database is a pivotal feature that enables developers to create applications capable of updating and syncing data instantly across multiple devices This functionality is especially beneficial for developing chat applications, social media platforms, and collaborative tools.
Firebase offers a range of authentication methods, including email and password, as well as social logins through Google, Facebook, and Twitter This enables developers to seamlessly integrate secure user authentication and authorization into their applications.
Firebase empowers developers to efficiently store and retrieve data in the cloud, facilitating app scalability and management of extensive data volumes Additionally, it offers hosting and cloud functions, simplifying the deployment and execution of serverless functions in the cloud.
Overall, Firebase is a powerful platform that provides a variety of tools and services to help developers build high-quality apps quickly and easily
Figure 2.9 Example Real-time Database
Firebase Realtime Database is a NoSQL cloud-hosted database that enables developers to create real-time applications, facilitating data synchronization across multiple clients instantly As a key feature of Firebase, it is extensively utilized for developing applications like chat platforms, collaborative tools, and social media networks.
The Realtime Database employs a JSON data model, enabling developers to store and synchronize data in real-time across clients This functionality ensures that any updates to the database are instantly reflected to all connected clients, guaranteeing they all access the most current data.
The Realtime Database offers robust features and APIs for effective data querying, filtering, and managing security and access control Developers can seamlessly integrate the Realtime Database into their applications using Firebase's SDKs for Android, iOS, and the web.
The Realtime Database offers significant scalability, capable of managing extensive data while supporting millions of users and thousands of simultaneous connections Furthermore, Firebase equips developers with comprehensive tools for monitoring and debugging the Realtime Database, facilitating the identification and resolution of issues efficiently.
The Firebase Real-time Database is a versatile and efficient solution for developers looking to create real-time applications swiftly Its real-time synchronization, advanced querying and filtering options, and strong scalability make it a favored choice for building dynamic applications.
INTRODUCTION WEBSITE
HTML, CSS, and JavaScript are essential front-end technologies that shape the user interface and experience of websites They collaborate with back-end technologies, including databases, servers, and APIs, to develop comprehensive web applications.
HTML forms the backbone of a web page's structure and content, while CSS enhances its visual appeal and layout By utilizing CSS, developers can manage colors, fonts, spacing, and various visual components, enabling the creation of responsive designs that adjust seamlessly to different screen sizes and devices.
Figure 2.10 Illustrated examples of HTML, CSS and Javscript
JavaScript enhances web pages by introducing interactivity and dynamic features like drop-down menus, slideshows, and animations It effectively manages user interactions, including clicks and mouse movements, while also facilitating real-time data updates through communication with back-end servers and APIs.
Front-end technologies are essential to modern web development, continuously evolving to address the dynamic demands of the internet Developers leverage various tools and frameworks, including React and Angular for JavaScript, as well as Bootstrap and Foundation for CSS, to enhance their projects effectively.
Web developers utilize both front-end and back-end technologies to enhance the functionality of web applications Back-end technologies, including databases, servers, and APIs, play a crucial role in this process Databases store and retrieve data, servers manage client requests and deliver content, while APIs facilitate communication between different systems and access external data sources.
Web development is a dynamic and multifaceted field that demands a wide range of skills and expertise At its core, modern web development relies on HTML, CSS, and JavaScript, which serve as foundational technologies However, these languages are only a component of a broader web development ecosystem that encompasses back-end technologies, various frameworks, essential tools, and industry best practices.
An Application Programming Interface (API) is a collection of protocols, routines, and tools that facilitate communication between software applications By enabling different applications to interact, APIs allow for data exchange and the execution of various tasks, enhancing functionality and integration across platforms.
APIs play a crucial role in websites by allowing third-party developers to access data and functionalities from web applications For instance, an e-commerce site may provide an API that shares pricing and product details, facilitating the development of new applications that utilize the e-commerce platform's data and features.
APIs can be utilized through various programming languages and protocols, including REST (Representational State Transfer) and SOAP (Simple Object Access Protocol) Among these, RESTful APIs have gained immense popularity in web development, primarily because of their simplicity and scalability.
APIs play a crucial role in web development by facilitating communication between diverse software applications, allowing for data exchange and task execution They provide third-party developers with access to a website or web application's data and functionalities, and can be utilized through various programming languages and protocols.
INTRODUCTION APP WITH ANDROID STUDIO
Android Studio is a widely-used integrated development environment (IDE) for creating mobile applications on the Android operating system It enables developers to build a diverse array of apps, ranging from basic applications to sophisticated ones with advanced functionalities and services.
Android Studio offers a comprehensive suite of tools and features designed to assist developers in creating high-quality, performance-driven mobile applications Its design editor enables real-time creation and previewing of app layouts, facilitating the development of visually attractive and user-friendly interfaces.
Android Studio offers an extensive array of APIs and libraries, facilitating seamless integration with Google services like Maps, Firebase, and Google Cloud Platform This capability enables developers to build robust and feature-rich applications effortlessly.
Figure 2.12 Example of an App Mobile
Android Studio prominently supports Kotlin, a modern programming language known for its conciseness, expressiveness, and safety compared to Java As Kotlin gains popularity among Android developers, its full integration into Android Studio enhances the development experience.
Android Studio offers an extensive suite of testing and debugging tools for mobile app development, featuring a built-in emulator that enables developers to test their applications on a virtual Android device This functionality ensures that the app operates correctly and is free of bugs prior to its public release.
Android Studio is a robust and flexible development environment ideal for creating mobile applications It offers a wide range of tools and features, supports Kotlin programming, and benefits from a vast community of developers, making it the preferred IDE for Android app development.
INTRODUCTION POWER BI
Power BI, developed by Microsoft, is a powerful business and data analytics service that allows professionals to effectively analyze, process, and visualize extensive data sets This platform facilitates the extraction of valuable insights, enabling users to make informed decisions and share interactive reports and dashboards across various departments.
Power BI enables users to effortlessly generate reports and dashboards through its user-friendly drag-and-drop interface, complemented by a diverse selection of interactive data visualizations This functionality facilitates the development of visually engaging and insightful reports that effectively convey intricate data.
Power BI is an effective tool that empowers businesses to make informed, data-driven decisions by delivering actionable insights in an intuitive format With its user-friendly interface and extensive features, it is the perfect solution for professionals aiming to enhance their data analysis and visualization workflows.
Power BI is faster and performs better when there is a smaller volume of data
Power BI provides an interface based on Microsoft Office 365 that is user-friendly, intuitive, and easy to understand
Using Power BI, you can work with several data sources, such as Excel, Text/CSV, JSON, SQL Server databases, IBM DB2, MySQL, etc
Power BI can connect with the R programming language, and it also supports various Data Analysis Expression (DAX) functions and measures
Power BI has functional integration with the Microsoft Azure cloud platform It helps to analyze insights and patterns in datasets
Figure 2.13 Example of Power BI
A Power BI dashboard is a concise, single-page document that visually narrates data stories through charts, graphs, and various visualizations Its primary goal is to present crucial information and insights clearly and quickly, offering a high-level overview of the data due to its one-page limitation.
If more detailed information is required, related reports can be accessed
Power BI stands out as the sole data analytics service that provides dashboards, often referred to as a "canvas" for single-page visualizations It's important to highlight that these dashboards can exclusively be created within the Power BI Service, not in Power BI Desktop.
Power BI dashboards serve as essential tools for businesses and professionals aiming to efficiently convey insights and data visualizations to stakeholders With thoughtfully crafted dashboards, users can swiftly recognize significant trends, patterns, and insights, enabling them to make informed, data-driven decisions.
INTRODUCTION TABLEAU
Tableau is a leading data analysis and visualization tool that allows users to create dynamic and interactive visuals without coding knowledge It enables the swift transformation of raw data into insightful visualizations, facilitating informed decision-making.
Tableau's standout advantage lies in its capacity to efficiently handle vast datasets and generate visualizations quickly, making it an essential tool for businesses and professionals seeking to swiftly analyze and convey insights from their data.
Tableau provides an extensive array of features, including an intuitive drag-and-drop interface, real-time collaboration capabilities, and seamless integration with various data sources It supports multiple data formats, such as Excel, CSV, and SQL databases, making it a versatile tool for data visualization and analysis.
Tableau is an effective data analysis and visualization tool that enables businesses and professionals to transform raw data into actionable insights With its user-friendly interface, rapid processing capabilities, and extensive features, Tableau is a favored option for data-driven organizations aiming to achieve a competitive advantage.
DESIGN AND IMPLEMENTATION
HARDWARE DESIGN
Python is a widely favored programming language for face recognition due to its simplicity and flexibility, complemented by robust libraries Among these, OpenCV stands out as one of the most popular tools for computer vision and face recognition tasks.
OpenCV offers an extensive array of tools and functions for effective image and video processing, particularly in the realm of face detection and recognition It enables developers to seamlessly implement advanced algorithms, including Eigenfaces, Fisherfaces, and Local Binary Pattern Histograms, which are widely utilized in various face recognition applications.
Python's simplicity and readability make it an ideal choice for face recognition projects, enabling rapid prototyping and experimentation With a large and active community of developers, there is an abundance of resources and support available for those working in this field.
Python is a versatile programming language suitable for developing face recognition systems across multiple platforms, including desktop, web, and mobile applications, allowing for deployment in diverse environments.
In summary, Python stands out as a top choice for face recognition due to its user-friendly nature, flexibility, and robust libraries It enables developers to seamlessly implement sophisticated algorithms while offering extensive resources and community support Additionally, Python's versatility makes it perfect for creating face recognition systems suitable for diverse applications.
First, we need python 3.8 and OpenCV:
Also we need the Dlib Package:
Figure 3.3 Installing the Dlib Package
Figure 3.4 Installing Face_recognition Package
Export library and area face:
Figure 3.5 Export library and area face
Next, encode images from the database
Figure 3.6 Encoding images from database
Use the camera by using cv2 and create frame that cover the face:
Figure 3.7 Create frame that cover the face
Compare the face have on database and face recognize detected, and calculate the distance:
Users can manage multiple devices within the system, including turning lights on and off, adjusting temperature settings, and opening or closing doors, through an intuitive interface on a website or mobile app Control commands are transmitted to the ESP32 using Firebase and RF21L01, enabling the devices to respond accordingly.
The system provides timely alerts to users when critical parameters like temperature, humidity, or air quality surpass safe limits These notifications are accessible through the website or mobile app interface and are sent directly to users via email or text message.
Parameter monitoring: Users can monitor parameters such as temperature, humidity, light, and air quality of the indoor environment through the I2C protocol and send these parameters to Firebase for transmission
3.1.2.1 Block diagram of face recognize
Figure 3.10 Block diagram of face recognize
The block diagram of the system includes the following main blocks:
Input image pre-processing is essential for optimizing the recognition process It involves several crucial steps, including brightness normalization, resizing the image to a standard dimension, and rotating the image to enhance recognition accuracy.
Feature extraction: This block is used to extract facial features from the input image
The features extracted include unique points such as nose, eyes, mouth, etc., which are used to determine the unique characteristics of each face
The classifier block is essential for comparing the extracted facial features against those stored in the face database, enabling the classification process to determine if a face is already registered in the system.
The Face Database serves as a repository for storing registered faces by capturing and preserving their unique extracted features These facial features are systematically extracted and securely stored within the database for efficient retrieval and analysis.
The face recognition system, developed in Python with OpenCV, leverages advanced image processing and computer vision tools Utilizing machine learning algorithms, it accurately recognizes faces by analyzing extracted features This versatile system is applicable in numerous fields, including access control, attendance monitoring, and security surveillance.
3.1.2.2 Block diagram of smart warehouse system
Our system employs two nodes that wirelessly communicate through RF24L01 technology, strategically positioned in the operation house and the warehouse Each location is outfitted with sensors and control devices to enhance functionality Node 1 acts as the root node and features a control panel for Node 2, ensuring effective communication even without Wi-Fi connectivity.
Fi Additionally, Node 1 facilitates communication with the app and Firebase
Figure 3.11 Interaction between Node 1 and Node 2
Figure 3.12 Block diagram Node 1 ( Operation room)
Figure 3.13 Block diagram Node 2 ( Warehouse Room)
The ESP32, developed by Espressif Systems, is a versatile and cost-effective microcontroller that integrates Wi-Fi, Bluetooth, and BLE capabilities Its low-power design makes it ideal for various applications.
The ESP32, a versatile microcontroller developed by Espressif Systems, is highly regarded for its capabilities in IoT application development Known for its robust features, it plays a crucial role in enhancing wireless communication technologies.
ESP-Wroom-32 contains a low-power Tensilica Xtensa® Dual-Core 32-bit LX6 microprocessor at 240 MHz: 994.26 CoreMark; 4.14 CoreMark/MHz
448 KB of ROM for booting and core functions
520 KB of on-chip SRAM for data and instructions
Bluetooth v4.2 BR/EDR and Bluetooth LE specifications
The ESP-WROOM-32 development board boasts a powerful Tensilica Xtensa® Dual-Core 32-bit LX6 microprocessor, capable of operating at adjustable clock frequencies ranging from 80 to 240 MHz It is equipped with 448 KB of ROM, 520 KB of on-chip SRAM, and 4MB of Flash Memory, making it an ideal choice for various embedded applications.
SOFTWARE DESIGN
The software development aims to create an application that monitors parameters, visualizes data, and enables online control of the hardware system To enhance security and establish ownership, the software will include a login page.
With the above purpose, building a website or IoT application, the software functions may include:
Collecting data from sensors and other measuring devices
Control operations on the website using voice commands
Storing data on a cloud database like Firebase for easy access from anywhere
Displaying data in the form of charts and graphs for easy observation and analysis
Providing device control features such as on/off, brightness increase/decrease, temperature, humidity, etc
Being able to interact with other smart devices in the system
Integrating notification features to alert when there is a malfunction or predefined limit is exceeded
Being able to update and manage the software remotely to increase the stability and reliability of the system
To enhance its functionality, the website has incorporated Power BI for effective management of sales, revenue, profit, and expenses, facilitating data visualization and statistical analysis for accurate revenue forecasting Additionally, the integration of Tableau allows for the monitoring of weather data across provinces and cities in Vietnam, enabling precise forecasting of future weather patterns.
Power BI enables in-depth data analysis on websites, delivering valuable insights into sales trends and revenue performance This analytical capability empowers businesses to make informed decisions regarding their product offerings, marketing strategies, and pricing approaches.
The Tableau integration allows real-time monitoring of weather patterns and access to historical data, enhancing logistics and transportation planning This valuable information offers insights beneficial to industries such as agriculture and tourism.
Overall, the added functionalities of Power BI and Tableau provide valuable insights that can help businesses make better decisions, optimize operations, and improve the customer experience
Here are the detailed functionalities of the blocks in the software block diagram for the IoT website:
User interface block: Represents components related to the user interface, including home page, login page, data display pages, and device control pages
Processing block: Represents components that process data and perform software functions such as data collection from sensors, data storage, data processing, and data management
Connection block: Represents components that connect to devices, servers, or other network services to transmit and receive data
Security block: Represents security components such as login and user authentication, access control, data encryption, and security monitoring
Online block: Represents components that monitor and control devices online, including control and status update features
Visual data block: Represents components used to display parameters and visualize data
Integration block: Represents components that integrate with data analysis tools such as Power BI and Tableau to help visualize data, statistics, and predict revenue
Figure 3.40 Visual Data Block Diagram
Log in to Firebase and create a project:
Figure 3.41 Log in to Firebase and create a project
Select Realtime Database and create a new database:
Figure 3.42 Select Realtime Database and create a new database
Create variables for reading sensor values within the Data on Board:
Figure 3.43 Create variables for reading sensor values within the Data on Board
Create variables for transmitting control signals to the hardware:
Figure 3.44 Create variables for transmitting control signals to the hardware
Figure 3.46 Setting up voice control
Setting control signal parameters for device control:
Figure 3.48 Setting control signal parameters for device control
Create a new project on Android Studio, Select "Empty Activity" and click "Next" to continue :
Figure 3.49 Create a new project on Android Studio
After creating the app, proceed to connect the app to Firebase:
Go to the Firebase website and create a new project
Add a new Android app to the project and enter the package name of your app
Download the "google-services.json" file and add it to the "app" folder in your Android Studio project
Add the Firebase SDK to your app by following the instructions provided by
Initialize Firebase in your app by adding the necessary code to your app's
Test the connection by running your app and checking the Firebase console for any data updates
Figure 3.50 Create a new project on Android Studio
To design the app interface in the activity_main.xml file:
Open the activity_main.xml file in the "res/layout" folder of your Android
Use the "Design" tab to drag and drop UI components such as text views, buttons, and image views onto the design canvas
Use the "Attributes" tab to customize the properties of each UI component, such as text size, color, and layout
Use the "Code" tab to edit the XML code directly if needed
Preview the app interface by using the "Preview" tab or by running the app on an emulator or device
Continue to refine and adjust the UI design until it meets your requirements and looks visually appealing.
RESULTS AND ASSESSMENT
HARDWARE
Evaluation and Expansion of Face Recognition:
The face recognition system demonstrated a recognition rate of approximately 70-80% for data stored in its database In contrast, the performance declined notably for external data, with up to 80% of cases incorrectly identified.
Despite the ability to recognize multiple faces at once, factors like poor lighting and low camera quality can lead to a 50% false recognition rate The system relies on complete facial landmarks—such as the eyes, nose, and mouth—for accurate identification, which poses challenges in recognizing individuals who wear accessories like masks, hats, or sunglasses.
Figure 4.1 Result of face recognition
The face recognition system effectively identifies 3-4 individuals at once from its database; however, external factors can significantly increase the false recognition rate, potentially reaching up to 50%.
We utilized SolidWorks design software to create the framework of the house This software provided us with simple tools that facilitated a quick and efficient completion of the project
Figure 4.2 Create the framework of the house with SolidWorks
We have created an innovative smart warehouse model featuring two distinct areas: an operation house and a storage room Each space is outfitted with advanced sensors and control devices, while the operation house includes an LCD screen and a button system for effective monitoring and control The two rooms maintain seamless wireless communication through the NRF24L01 module, utilizing radio wave technology for efficient data exchange.
Sensors and devices system in every house includes : DHT 11, Gas (MQ4), Light sensor, Fan, Light
Figure 4.4 Sensors and devices system
- LCD will be displayed control parameters: fan1, fan2, light1 , light2, set, mode
In our system, we have different operating modes, where 1 represents the 'ON' state and
The system features a mode-setting option where mode=1 activates manual control through physical buttons or web/app interfaces, while mode=0 signifies the 'OFF' state and activates automatic mode based on pre-programmed conditions.
In manual control mode (mode=1), when set=H, the system operates through hardware buttons, and when set=W, the control is performed using software through web or app interfaces
After analyzing the operational patterns, we discovered that the communication data exchanged between the two rooms using the NRF24L01 wireless module remains relatively stable Continuous transmission tests revealed that data is consistently received, with delays ranging from 0.5 to 0.75 seconds, influenced by distance and potential obstructions.
The system's buttons utilize a low-level active circuit design, ensuring maximum safety for the processor by preventing over-current and over-voltage issues Observations during extended model operation indicate a button press delay of approximately 0.3-0.5 seconds, allowing for rapid updates to the button status.
The LCD 16x2, while compact, effectively displays essential system parameters for controlling lights and fans By updating every three seconds, it minimizes the program's workload while ensuring reliable display performance.
In our program, we utilize various sensors, including the DHT11 temperature and humidity sensor, a light sensor, and a gas sensor Our experiments indicate that the temperature and humidity readings achieve an accuracy of approximately 95% over short periods Notably, the light sensor displays a decrease in analog value with increasing light intensity, with values around 425 indicating strong light Additionally, the gas sensor registers an analog value exceeding 1000 when an igniter is brought close, signaling the presence of gas.
Controlling hardware through web and app interfaces via Firebase typically incurs a delay of approximately 3 to 5 seconds, influenced by internet connectivity This level of latency is generally acceptable for systems that do not require immediate responsiveness.
The system demonstrates stable performance with low latency; however, its sensor accuracy remains limited due to the use of standard sensors and the impact of environmental factors and location.
These buttons, when pressed, generate a low-level signal that the microprocessor detects and activates, after which it returns to a high-level state
The indoor monitoring, control, and warning system utilizing IoT technology features components such as BUTTON, LCD, DHT11, MQ4, and LIGHT SENSOR, all demonstrating high accuracy, stability, and low error rates The BUTTON components ensure reliable device control with minimal latency, promoting seamless user interactions Additionally, the NRF24L01 communication protocol enables stable and reliable wireless communication between separate rooms without the need for physical connections.
The DHT11 and MQ4 sensors offer exceptional accuracy and stability in monitoring temperature, humidity, and air quality, ensuring users receive trustworthy insights into their indoor environment Additionally, the LIGHT SENSOR effectively delivers precise measurements of indoor lighting conditions.
The BUTTON components have been reliable in controlling devices, with low latency and stable operation Users can control their devices remotely with ease, adjusting settings according to their preferences
The NRF24L01 communication protocol facilitates reliable wireless communication across different rooms, removing the necessity for physical connections Our testing confirms that the connection remains stable, allowing users to seamlessly access and monitor their indoor environment without interruptions.
The successful implementation of an indoor monitoring, control, and warning system using IoT technology with ESP32 hardware enables remote management of devices like lights, fans, and air conditioners Utilizing the RF21L01 communication protocol ensures reliable data transmission, providing users with accurate and timely information to effectively adjust their indoor environment.
WEBSITE AND APP MOBILE
The indoor monitoring, control, and warning system utilizing IoT technology has demonstrated exceptional effectiveness in its data visualization feature on both the website and app By accurately reading sensor data from Firebase, the system offers real-time updates, enabling users to stay informed about their indoor environment.
The website and app offer an intuitive data visualization feature, showcasing various charts and graphs that enable users to effortlessly monitor changes in their indoor environment over time This system presents crucial information on temperature, humidity, light levels, and gas concentrations, delivering a comprehensive overview of indoor conditions.
Figure 4.7 Sensor parameters from Firebase
The highly customizable charts and graphs enable users to tailor their data visualization according to personal preferences, including selecting specific time frames, chart types, and data parameters This extensive customization ensures quick and easy access to essential information.
Figure 4.8 Data visualization with a Website
Figure 4.9 Data visualization with App Mobile
The data visualization feature of the IoT-based indoor monitoring, control, and warning system is highly effective, providing real-time sensor data on the website Users can easily track changes in their indoor environment through intuitive charts and graphs Additionally, the system offers a high level of customization, allowing users to quickly and easily access the information they need.
The indoor monitoring, control, and warning system leverages IoT technology to enable remote device management via user-friendly buttons on a website These buttons send signals through Firebase, allowing the hardware to interpret the commands and perform the desired actions efficiently.
This feature allows users to effortlessly control their devices remotely, eliminating the need for physical interaction By simply clicking the buttons on the website, users can easily adjust their device settings, making it a highly user-friendly and accessible option.
Figure 4.10 Signal control from Firebase
The website offers voice control integration alongside traditional BUTTON control, allowing users to operate their devices through voice commands This functionality enhances accessibility, especially for individuals with mobility challenges or in scenarios where physical interaction is impractical.
Figure 4.11 Control Device with Website
The integration of BUTTON and voice control features in the IoT-based indoor monitoring, control, and warning system represents a groundbreaking advancement in device management BUTTONS offer users a straightforward and convenient method for remote device control, while the addition of voice control enhances accessibility and ease of use These innovative features highlight the system's dedication to delivering a comprehensive and user-friendly solution for indoor monitoring and control.
4.2.3 STORE MANAGER WITH POWER BI
This study aims to evaluate the effectiveness of the Store Manager utilizing the Power BI tool in overseeing sales, profits, and costs within a retail environment Efficient management of these elements is crucial for the success of retail businesses The Store Manager, equipped with the Power BI tool, offers a comprehensive and intuitive solution for analyzing and managing these key factors.
In this study, we analyzed sales, profits, and cost data for various retail items using the Store Manager with Power BI tool This allowed us to visualize the data through charts, graphs, and tables, while also generating precise forecasts tailored to user-specific needs by adjusting various parameters.
Our analysis revealed that the Store Manager utilizing the Power BI tool significantly enhanced the management of sales, profits, and costs in retail operations This comprehensive and user-friendly solution enabled businesses to make informed operational decisions by presenting data in an accessible format, including charts, graphs, and tables Additionally, the tool's advanced analytical capabilities facilitated precise forecasting of sales, profits, and costs, further optimizing retail performance.
Figure 4.12 Managing a store with Power BI
The Store Manager with Power BI is an essential solution for retail businesses aiming to optimize sales, profits, and costs Its user-friendly interface and advanced analytical features empower businesses to analyze key performance metrics effectively By leveraging this tool, retailers can make informed decisions that enhance operational efficiency and profitability, ensuring they remain competitive in the market.
This study aims to analyze and predict weather patterns in Vietnam utilizing the Forecast Weather with Tableau tool Accurate weather forecasting is essential for various sectors, such as agriculture, transportation, and tourism By delivering real-time and precise weather data, the Forecast Weather with Tableau tool enables businesses and individuals to make well-informed decisions regarding their plans and activities.
In this study, we gathered and analyzed historical weather data from multiple locations in Vietnam, utilizing the Forecast Weather with Tableau tool to produce precise weather forecasts tailored to specific needs To enhance user understanding and analysis, we presented the data through various formats, including maps, charts, and tables.
Our analysis revealed that the Forecast Weather with Tableau tool is highly effective for generating accurate weather forecasts across Vietnam It offers real-time, up-to-date weather information, enabling users to make informed decisions regarding their activities and plans The data is presented in an accessible format, featuring maps, charts, and tables, which enhances user-friendliness.
Figure 4.13 Weather forecasting with Tableau Conclusion:
CONCLUSION AND RECOMMENDATION
CONCLUSION
We have developed an advanced indoor monitoring, control, and warning system leveraging IoT technology, ensuring secure and convenient access to various devices and sensors within indoor environments This innovative system employs the ESP32 microcontroller along with the RF21L01 communication protocol for seamless wireless communication Additionally, it incorporates facial recognition technology to enhance door access security.
- and Firebase for connecting the system with web and mobile applications
The project has achieved its objectives of providing a user-friendly interface to control and monitor devices in an indoor environment, and also predicting weather forecasts and visualizing sales data.
COMPLETED WORKS
The completed works of this project include:
- Design and implementation of the indoor monitoring, control and warning system using IoT technology
- Integration of facial recognition technology for door access
- Integration of Firebase for cloud-based data storage and connection with web and mobile applications, control and monitor equipment, monitor sensor data
- Implementation of a web-based dashboard for data visualization and sales forecasting
- Implementation of a weather forecasting system
- Testing and evaluation of the system's performance and reliability.
LIMITATIONS
The smart warehouse system operates using two nodes located in separate rooms, which communicate wirelessly via the NRF24L01 module This module has a communication range of less than 100 meters, provided there are minimal obstacles Although it supports a packet size of 32 bytes, this may be insufficient for larger data requirements in extensive systems Additionally, the absence of a feedback mechanism for error transmission necessitates continuous data sending In a multi-node configuration, managing address division and transmission line identification becomes complex Furthermore, while the NRF24L01 enables fast data transfer rates, this can compromise reliability over longer distances.
The system relies on outdated sensors, resulting in lower data reliability and accuracy compared to real-world conditions Additionally, the limited number of sensors contributes to a lack of data diversity.
The system operates through a mobile app and web platform, utilizing Firebase for communication However, this reliance means the application becomes inactive when the desktop is powered off Additionally, the system lacks a primary Wi-Fi configuration feature, resulting in signal loss or disruption when Wi-Fi is weak or unavailable, leading to a default automatic operation mode.
- The system is limited to indoor environments, and cannot be used in outdoor environments
- The facial recognition technology may not be 100% accurate, and may fail to recognize some users
- The system depends on a stable and reliable internet connection, which may not always be available
- The RF21L01 communication protocol has limited range, and may not be suitable for larger indoor environments.
IMPROVEMENT SUGGESTION
To address the limitations of the system, the following improvement suggestions are recommended:
- Develop an outdoor version of the system to extend its application to outdoor environments
- Improve the facial recognition technology to enhance its accuracy and reliability
- Develop a backup system to ensure the system can continue to function in the event of internet connection failure
- Use a more powerful communication protocol to improve the range of the system in larger indoor environments
In summary, the IoT-based indoor monitoring, control, and warning system offers a valuable solution for boosting security and optimizing energy efficiency in indoor spaces With ongoing advancements, this system has the potential for broader applications across various environments.
Improvement details for face recognition:
- Enhance the recognition algorithm to improve accuracy for external data and reduce false recognition
- Implement advanced image processing techniques to handle challenging lighting conditions and camera quality
- Investigate and integrate additional biometric features (e.g., voice recognition) to enhance overall identification reliability
- Explore deep learning models for improved facial feature extraction and recognition performance
- Develop a comprehensive training dataset with diverse facial variations and accessories to boost the system robustness
- Employ 3D facial recognition to capture more comprehensive facial information and handle occlusions effectively
By focusing on these key elements, we strive to enhance the face recognition system, delivering a more dependable and adaptable solution for practical applications, while significantly increasing accuracy and reducing false recognition rates.
[1] arrow.com “What is I2C? How Inter-Integrated Circuits Work“ Available: https://www.arrow.com/en/research-and-events/articles/what-is-i2c-how-inter-integrated- circuits-work, February 4, 2019
[2] analog.com “Introduction to SPI Interface” Available: https://www.analog.com/en/analog-dialogue/articles/introduction-to-spi-interface.html,
[3] techtarget.com “face detection” Available: https://www.techtarget.com/searchenterpriseai/definition/face- detection#:~:texte%20detection%20software%20detects%20faces,it%20has%20detect ed%20a%20face, April 2023
[4] blog.vietnamlab.vn “Nhận diện khuôn mặt với Python và OpenCV” Available: https://blog.vietnamlab.vn/nhan-dien-khuon-mat-voi-python-va-opencv/ , June 29, 2018
[5] academia.edu “Smart Home Systems: Overview and Comparative Analysis” Available: https://www.academia.edu/41431240/Smart_Home_Systems_Overview_and_Comparative_ Analysis, 2018
[6] icdayroi.com “kít RF thu phát wifi BLE ESP32 NodeLuaNode32 38 chân” Available: icdayroi.com/kit-rf-thu-phat-wifi-ble-esp32-nodemcu-luanode32-38-chan
[7] thegioiic.com “NRF24L01 + PA + LNA AI-Thinker Mạch Thu Phát RF 2.4Ghz
1000m” Available: https://www.thegioiic.com/nrf24l01-pa-lna-ai-thinker-mach-thu-phat-rf- 2-4ghz-1000m
[8] thegioiic.com “DHT11 Mạch Cảm Biến Nhiệt Độ Và Độ Ẩm Đen” Available: https://www.thegioiic.com/dht11-mach-cam-bien-nhiet-do-va-do-am-den
[9] nshopvn.com “Cảm biến khí Metan MQ-4” Available: https://nshopvn.com/product/cam-bien-khi-metan-mq-4/
[10] thegioiic.com “LM393 Cảm Biến Ánh Sáng Quang Trở 3 Chân” Available: www.thegioiic.com/lm393-cam-bien-anh-sang-quang-tro-3-chan
[11] thegioiic.com “LCD 1602 Nền Xanh Lá Chữ Đen 5V Kèm I2C Driver” Available: www.thegioiic.com/lcd-1602-nen-xanh-la-chu-den-5v-kem-i2c-driver
[12] how2electronics.com “STM32 NRF24L01 Sensor Node with ESP32 NRF24L01
Gateway”.Available: https://how2electronics.com/stm32-nrf24l01-node-with-esp32- nrf24l01-gateway/, May 29, 2023
[13] randomnerdtutorials.com “How to Use I2C LCD with ESP32 on Arduino IDE
(ESP8266 compatible)” Available: https://randomnerdtutorials.com/esp32-esp8266-i2c-lcd- arduino-ide/
[15] randomnerdtutorials.com “Getting Started with the ESP32 Development Board”
Available: randomnerdtutorials.com/getting-started-with-esp32/