Chapter 1

An IoT System for Earthquake Detection

By Yee-Loo Foo [0000-0001-6106-4655] and Ahmed Bakr Alagha [0000-0002-3228-4576]

Abstract – Earthquakes can strike at any time, posing a threat to human lives. Over the last decade, there has been an increase in both the frequency and severity of earthquakes. It becomes critical to have a fast and accurate earthquake detection system that is affordable to the general public. We have constructed one using some low-cost, off-the-shelf IoT devices. To be specific, we have used the ADXL345 accelerometer as the seismic sensor. An ESP32 microcontroller controls the ADXL345 accelerometer and a GPS module NEO-6M. When an earthquake is simulated, the ADXL345 measures the amount of acceleration and reports it to the ESP32 microcontroller. The GPS module then gathers the location information in terms of latitude and longitude. A LCD display reveals the magnitude of the earthquake to the user. On the other hand, a smartphone app has been developed in order to alert the users on potential earthquakes, and also to guide the users on the cautious steps to be taken when there is an earthquake. The sensing data is collected and stored in a SQLite database. A Django server hosts the web application, and the smartphone app is developed using Flutter.

Chapter 2

Structural Health Monitoring Using IoT

By Yee-Loo Foo [0000-0001-6106-4655] and Punitha Shanmugam [0000-0003-0720-9563]

Abstract– Destructions of manmade structures are caused by various reasons, e.g., catastrophes such as earthquakes and floods, overloading, etc. Close and effective monitoring of structural health is crucial. Issuing early warnings of unsafe structures may save many lives and properties. With the advancement of the Internet-of-Things (IoT) technology, it has become possible to implement a structural health monitoring (SHM) system at a low cost. We have devised and implemented a simple prototype using off-the-shelf products, e.g., sensors, single-board computers, etc. In brief, an ESP8266 microcontroller is used to control a vibration sensor SW-420, a Flex sensor (to measure the amount of bending), and a DHT-11 temperature sensor. The system connects to an IoT platform called Blynk. Users can access the data via the Blynk mobile app. Experimentations with the prototype provide us with some insights on the next step forward.

Chapter 3

Development for an IoT-based Smart indoor Home Gardening System

Hammad Anas Edris Moh, Azwan Mahmud [0000-0001-5281-2877], Azlan Abd Aziz [0000-0003-2255-0831], Mohamad Yusoff Alias [0000-0002-0766-7911], Hafizal Mohamad [0000-0003-1915-6710], and Syamsuri Yaakob [0000-0003-2988-6362]

Abstract – The growth of IoT technologies allowed devices all around the world to be connected, as well as to be monitored and controlled in real time. Such a technology can be used to automate tasks and reduce the amount of work done by humans. The system proposed here makes use of the advancements offered in the IoT technologies to automate the process of caring after indoor plants. The system consists of multiple sensors to monitor the different parameters of the plant such as soil moisture, light exposure, water level, temperature and humidity; As well as actuators such as a water-pump, and a light-bulb to provide the plant with water and light. The system has its own custom made IoT platform using Django that receives data from the microcontroller, process it, and decide whether to turn on/off any of the actuators based on the readings of the sensors. Furthermore, the system comes with a cross-platform mobile application designed using React Native which consists of three main pages. The first page is meant to keep the user updated about the status of the plant, obtained from the sensors. The second page is the control page which allows the user to control the actuators. The third page is to show historical data of the sensors in graphical view. Finally, notifications will be sent to the user through the application when certain events occur, such as if the water level in the thank is too low, or if the tank leaks water.

Chapter 4

IoT-Based Beehive Monitoring System for Stingless Beekeeping

By Nur Mimi Syaheeda binti Mohmad Faisal, Muhammad Irfan bin Mohd Pikri, Mohd Zulkifli Mustafa [0000-0002-8140-8184], and Siti Azlida Ibrahim [0000-0002-5909-4406]

Abstract – Stingless bee is a bee species widely found in tropical countries like Malaysia. It plays a vital role in conserving the natural ecosystem, biodiversity, and food security. One of the challenges in the stingless beekeeping industry is to monitor the status or the condition of the beehives. It is more difficult if the location of the hives is far from the residential area or if the number of beehives is large. It will be time-consuming for the beekeepers to visit the hive to check the conditions frequently. Therefore, technology with real-time remote monitoring capability will ease the burden on the beekeepers. In order to bring the stingless bee industry into a better stage and enable the establishment of large-scale apiaries, the application of technology such as the Internet of Things (IoT) is an excellent option to be explored. In this work, we propose an IoT-based monitoring system to monitor the beehive using sensors attached to the stingless beehive. The system counts the frequency of bees leaving and entering the hive, which is used to determine the status of the beehive. The system is connected to the internet via LoRa connectivity, which users can access using mobile apps or web browsers. By monitoring the beehive status, any sudden decrease in the colony can be detected, and the bee farmers can take proper action to avoid a more severe problem. The test results showed that the sensor data transmission was successful with a maximum distance of 200 meters between Lora node and the gateway, using a line-of-sight (LOS) connection. Meanwhile, the system achieves the maximum range of 60 meters for non-line-of-sight (NLOS) connection.

Chapter 5

Low-Code Reconfigurable IoT Sensor Nodes for Smart Agriculture

By Rizal Rusydi bin Hamka and Wooi Haw Tan [0000-0002-0436-0391]

Abstract – This project proposes a low-code IoT sensor node that can be reconfigured via a custom-made Arduino firmware and a graphical user interface (GUI) application. As an example to illustrate its usefulness, multiple sensor nodes are incorporated into a smart agriculture system for the purpose of reconfiguring an Arduino sensor’s threshold and behaviour, as well as to enable or disable Arduino’s actuators which act on the threshold. Several Arduino sensor nodes and a Raspberry Pi (RPi) gateway are fitted with Nordic Radio Frequency transceiver (nRF24L01+) modules to allow for wireless communication between them. An IoT platform, ThingsBoard, acts as a remote monitoring system that a user can view even outside of the farm. The Arduino communicates serially with the GUI which contain general details such as node name, radio address, and interval to update to ThingsBoard. The GUI allows for the reconfiguration of seven sensors – DHT11 for temperature/humidity, BH1750FVI for light reading, DS18B20 for soil temperature, FC-28 for soil moisture, MQ-2 for gas level, HC-SR501 for motion detection, and HC-SR04 for distance reading. The GUI settings along with the sensor readings get wirelessly communicated to the RPi, and the RPi sends these data to ThingsBoard for storage and display. The actuators are low-voltage fan, light bar, water pump, and two buzzers, and they are linked to the reconfiguration menu of appropriate sensors. In this project, the GUI has successfully stored user data into the Arduino’s Electrically Erasable Programmable Read-Only Memory (EEPROM) without explicit reprogramming, with the Arduino establishing wireless communication with the RPi, and the RPi pushing user data and sensor reading into ThingsBoard via Message Queuing Telemetry Transport (MQTT).

Chapter 6

Generalised Frequency Division Multiplexing (GFDM) Multicarrier Modulation for 5G and Beyond

Muhammad Nadzmi bin Mazlan and Ivan Ku [0000-0003-1405-7668]

Abstract – Wireless network technology has moved towards the fifth generation (5G) where the number of connecting devices is expected to increase by a hundred folds or more, resulting in a large volume of data traffic. Although the Orthogonal Frequency Division Multiplexing (OFDM) waveform is utilized in 4G to transmit data over the air, its high out-of-band emission (OOBE) makes it unsuitable to be adopted in 5G and beyond wireless networks where high data traffic densities are prevalent. The purpose of the research is to investigate alternative multicarrier modulation waveforms that are suitable for 5G and beyond wireless networks. The Generalized Frequency Division Multiplexing (GFDM) waveform is one such novel approach and will be analyzed here. The work here provides an independent analysis of its performance. Firstly, the system model for the generation of the GFDM waveform is presented. Then, the structure of the generated GFDM waveform is analyzed and its performance in terms of OOBE and supported data rates are evaluated through Matlab simulation and compared with the OFDM waveform. Simulation results show that the OOBE of GFDM can be up to 30 dB lower than the OFDM. Furthermore, its supported data rate improves by 18% as compared to OFDM when the carrier spacing is 15 kHz and the cyclic prefix duration is at 20% of one OFDM symbol period. These two improvements in GFDM implies that, for a given bandwidth, its waveform is able to carry more data over the air as compared to the OFDM waveform. This improves the bandwidth utilization efficiency which is crucial at scenarios with high data traffic densities.

Chapter 7

Sub-1V Pure CMOS Bandgap Reference Circuit using 0.13μm Technology

Shakti Kumaran Manugaran, Chu Liang Lee, Lini Lee [0000-0002-9686-5812] and Kah Yoong Chan

Abstract – A bandgap reference (BGR) is an important building block in integrated circuit design to provide an accurate voltage reference which is temperature insensitive, supply voltage independent and process variation independent. The use of BGRs is ubiquitous in applications that require high level of precision such as linear regulators, radio-frequency integrated circuit and many more. The conventional BGR voltage reference is 1.12V, however the trend towards smaller transistor and lower supply voltages has led to the demand of a lower voltage reference value in BGR design. For supply voltage as low as 1.20V, a sub-1V BGR with voltage reference lower than 1V is required. In this project, CMOS bandgap voltage reference output 484mV with low supply voltage of 1.20V is designed. This bandgap eliminated the used of parasitic vertical bipolar (BJT) transistor and purely implemented with CMOS transistors which allowed lower headroom for lower supply voltage circuit. This BGR circuit design is implemented using 0.13μm CMOS technology.

Chapter 8

Deep Learning Based Security System using Car Plate Information

Almaswari Osamah Abdullah Hezam, Gwo Chin Chung [0000-0002-3262-3451], It Ee Lee [0000-0002-0922-8859], and Khan Vun Teong [0000-0001-5139-620X]

Abstract – In our current era, the series of crimes in the community, the number of complexes and residential buildings is still a constant concern for residents, even with the presence of security services in buildings and surveillance cameras. Vehicle crimes are considered one of the highest crime rates, as, on the one hand, the crimes of vehicle theft are common when they are used as a means of committing crimes and robberies. Criminals take advantage of security loopholes, such as when a fake car licence plate is placed on the car, and the security system does not detect the fake car plate and allows the criminals to easily enter the building. The project’s goal is to detect a fake licence plate in a car using a deep neural network. The MobileNet SSD model, which is an architecture model of the convolution neural network (CNN), is used to predict vehicles, and the YOLOv4 model, which is a single-stage deep learning based object detector, is used to detect licence plates. The proposed system has successfully achieved results in detecting licence plates with the highest accuracy of 97% when placing surveillance cameras on the front sides of vehicles at a perpendicular angle and using high-resolution images. This automated detection process allows the security system to discover the use of fake car plate numbers and to prevent any crime from happening before it takes place.