This chapter is discussed about the
information that can be gathered to build up this project from the journal,
findings, article, and books. This chapter also give an explanation about the
related study done by other researchers. During literature review, the studies
will highlight few topics that will support and contributes the idea to make
this research reliable for those who studying in the same area. The research
that has been done by other researchers in fall detection system using wireless
sensors. This chapter will discuss the general information related in writing
to the project title which are fall detection system, wireless sensors, microcontroller,
GPS and GSM technology. Through the brief explanation in these topics, it will
help in better understanding of this project of what this project is all about.
Fall Detection System Overview
Falls are a common issue, but they are difficult to
define rigorously. Falls among the elderly is increasingly becoming a concern issue
in developing countries. The aftereffects of falls in elderly may lead to physical,
medical, psychological, social and economic consequences adversely. In
addition, a person who falls often lays for hours unable to move or use the alert
mechanisms. Sometimes, they are seriously injured or even dead by the time they
are found (Tang, Ong & Ahmad, 2015). Thus, researchers from all around the
world has invented various methods of fall detection in order to prevent these
fall detection system is an assistive device (Igual, Medrano, & Plaza, 2013)
which can detect the fall incidences and prevent the fall victims from lying
down for hours (Igual, et al., 2013) using motion-sensors devices. In early
years, a human fall detection system is intended to alert when a fall event
occurs but not to diminish the incidence of fallen (Dang,Truong & Dang, 2016).
This system can alert a concerned person and can also be easily build and use
2.1.1 Method of Fall Detection
technologies have been studied to detect falls, which can be separated in two
main approaches which are wearable devices and
context-aware systems, the sensors are installed in a certain area or
environment in order to detect falls event. There are various common type of
sensors which are floor sensors, pressure sensors, infrared sensors, cameras,
and microphones (Igual, et al., 2013).
An ambient device method is one of the subgroup in context-aware systems where
multiple sensors, most commonly pressure sensors are deployed in indoor
environment to detect the presence of user and when the user falls (Gannapathy, Tuani Ibrahim, Zakaria, Othman,
Latiff, 2013). According to Delahoz and Labrador (2014),
since the pressure sensors are low-cost and inconspicuous. The major benefit of
this system is the user need not to wear any device on their body (Arote & Bhosale, 2015).
Though, the major weakness of these sensors have is the low accuracy of fall
the other hand, wearable devices method often rely on small electronic devices which
must be worn by the user to be able to detect changes in acceleration, motion
or impact in order to detect falls event when it occurs and to immediately send
an alert to a concerned person. The user usually wear one or more wearable devices which are usually equipped with movement sensors such as accelerometers and
gyroscopes. The benefits of this method are not restricted to only a certain
area, low cost of installation and small in size
which will ease the user to perform his/her daily activities. Moreover,
according to a study by Wang,
Zhang, Bin, Lee, and Sherratt (2014), the accuracy of the system is very high
which 97.5% is while 96.8% of sensitivity and the specificity is 98.1%. In
addition, the user can be almost track universally if a wireless connection are
provided in the wearable devices (Casilari,
Luque, & Morón, 2015). Most of the existing human fall detection systems require
a base station. This proposed paper will develop a human fall detection system
which works as a wearable device, hence entirely not relying from a base
microcontroller is a micro, low-cost integrated computer, used for sensing
input from the real world and controlling devices based on that input. The
circuit contains many items that a desktop computer has such as memory and
Control Processor Unit (CPU) but does not have any interface such as monitor
and keyboard (Brain, n.d.).
easily to communicate with desktop computers and to use simple sensors and
output devices (Jayant,
n.d.). In addition, it is also very useful because it
does not need the full power of a desktop computer and does need to be smaller
or cheaper for designing a simple interactive device. Microcontroller has two
main types namely Arduino microcontroller and Raspberry Pi microcontroller.
2.2.1 Arduino Microcontroller
Uno as shown in Figure 2.1 is a single-board microcontroller and an open-source
platform based on ATmega328 (Tang, et al., 2015).
Arduino has an operating voltage of 5V while the input voltage range from 6V to
20V. It also has 32KB of flash memory, 2KB of SRAM, 1KB of EEPROM, 23
programmable I/O channels and up to 20MHz of clock. Arduino board has various functionalities including creating
interface circuits to read sensors and switches, writing programs and control
lights and motors with a slight effort. Moreover with using a USB cable, it can
load new code onto the board without needing a separate hardware which is also
called programmer. Arduino can be
programmed and controlled using Arduino Programming Language. It is also easier to learn to program it
since it uses a simplified version of C and C++ thus, do not need much
programming knowledge in order to do basic programming (Ramalingam, Dorairaj, & Ramamoorthy, n.d.).
However, there are still downsides of Arduino compared to Raspberry Pi
microcontroller. Firstly, it is not easy to connect to the internet. In
addition, Arduino is not great selection in writing complex software, entire
software stack or protocols for projects. Moreover, Arduino is not very
powerful compared to the Raspberry Pi.
2.1 Arduino Uno Board
(Source:Tang, et al., 2015)
Pi is a fully functional credit card-sized computer invented in year 2012,
originally designed especially for education because it is cheap and
education-oriented computer board. The raspberry pi is low cost, powerful and it
does not consume a lot of power (Kasundra & Shirsat, 2015).
Raspberry Pi is not only can be used as sensor node but also as a controller
where these can be used as processing node in WSN networks. In addition, the
processing of data and decision making can be based on artificial intelligence (Kasundra & Shirsat, 2015).
The technical specifications of Raspberry Pi 3 board will be simplified in
Raspberry Pi 3 Specifications
1GB LPDDR2 (900 MHz)
10/100 Ethernet, 2.4GHz 802.11n
Classic, Bluetooth Low Energy
1.2GHz Quad Core
Dual VideoCore IV
system or Windows 10 IoT
HDMI (rev 1.3
& 1.4), Composite RCA (PAL and NTSC)
3.5mm jack, HDMI, USB 4 x USB 2.0 Connector
15-pin MIPI Camera Serial Interface
Interface (DSI) 15 way flat flex cable connector with two data lanes and a
4 USB Ports
GPU has been capable of access, fast 3D core. The Raspberry Pi however,
does not have internal storage. However, it can use the SD cards to replace
internal storage thus, making the debug of the software updates or operating
system can be changed faster from different versions. Additionally, Raspberry
Pi runs entirely on open-source software and it is easy to install in most
Linux software since it runs a specially designed version of Linux OS.
ideal microcontroller for the proposed project is Raspberry Pi as shown in Figured
2.2 because the price is affordable in encouraging younger people for programming,
exploring, experimenting mastering, and inventing mainly for university
students. In addition, it was quickly adopted by inventers, makers, and
electronics enthusiasts for projects that require more than a basic
microcontroller such as Arduino devices because of its small size and
accessible price. Furthermore, Raspberry Pi can be connected to the internet
Figure 2.2 Raspberry Pi 3 Board
(Source: Qifan, Yang, Wang & Xu,
sensor defined as a device which responds or senses any parameter of physical
motion such as distance, rate and acceleration. Motion sensors may measure the changes
in pressure or magnetic field indirectly. The electronic sensor can be
integrated with or connected to other device where it measures motion which
contains an electronic sensor to alert the user of the presence of a moving
object (“What is A Motion Sensor?”, n.d.). These sensors includes gyroscope
sensor, accelerometer sensor, magnetic (Mag) Sensor and pressure sensor. In
this proposed project, gyroscope sensor and accelerometer sensor will be used
and will be explained more in further details.
2.3.1 Accelerometer Sensor
An accelerometer is
a low-power and an electromechanical
device that measures the vibration, or acceleration of motion of a structure (“A beginner’s
guide”, n.d.). It is a compressed device intended to measure
non-gravitational acceleration. The sensor
will respond to the vibrations related to movement such as from a firm movement
to any velocity when the object is integrated (Goodrich, 2013).
An accelerometer sense either static forces of acceleration which is
gravity or dynamic forces of acceleration which include movement and vibrations
(“Accelerometers Basics”, n.d.). Accelerometer sensor as shown in Figure 2.3 are
important components to devices that track fitness and other measurements in
the quantified self-movement. Accelerometers also allow the user to
understand the surroundings of an item better (Goodrich, 2013).
Figure 2.3 Accelerometer Sensor
2.3.2 Gyroscope Sensor
sensor, also known as angular velocity sensor is used for sensing and measuring
the angular velocity motion and changes in orientation of an object (Trusov, 2011; “Gyro sensors – How They Work and What’s Ahead,” n.d.).
A gyroscope manipulate gravity of the Earth and uses the key principles of
angular momentum in order to determine orientation (Goodrich, 2013).
The gyroscopes can measure velocity
of rotational in either one, two, or three axes. 3-axis gyroscopes as shown in
Figure 2.4 consists of a wheel where a rotor is mounted onto smaller spinning
axis which placed in the center of a more stable wheel (Goodrich, 2013).
There are three main types of gyroscope such as Rotary Gyroscope, Vibratory
Gyroscope and Optical Gyroscope.
Figure 2.4 3-axis Gyroscope Sensor
Positioning System (GPS) is as a satellite-based navigation system which uses
radio signals to determine the position precisely. It was originally for
military purposes only, however was made available for public use eventually
and it was developed by the US Department of Defense in 1973 but not fully
functional until 1994 (“How GPS Works”, 2014).
In short, it is mainly a navigation system for real-time positioning and being
used all over the world to determine coordinates and to navigate in air, on
land and sea accurately by GPS receiver. The GPS receiver as shown in Figure 2.5
analyzes the transmitted radio signals from the GPS satellites and measure the
time duration of the signals travelling from satellite to the receiver in order
to tracking the location of minimum of four GPS satellites and measure the
distance between each of GPS satellites and the receiver (“GPS in Schools – How
GPS Works”, 2015).
Figure 2.5 GPS Module
(Source: Borle & Kulkarni, 2016)
Global System for
(GSM) Modem is a wireless
modem that works with a GSM wireless network to transfer data. GSM modem
requires a SIM card from a wireless operator to enable it transfer data through
the operators’ network. GSM modem as shown in Figure 2.6 is controlled by a special
set of commands known as AT commands (“Introduction to AT Commands”, n.d.). SIM800
GSM has 68 Surface Mount Technology (SMT) pads and provides interfaces between
the module and all user’s hardware. It has GPRS multi-slot class 12/class 10,
One PWM, Bluetooth function, Audio channels, one SIM card interface and
operates on Quad-band 850/900/1800/1900 MHZ and transmit the SMS and Voice data
with low power consumption.
Figure 2.6 GSM Module
& Kumar, 2015)
The studies on human fall detection system has
boost up in recent years due to the advanced of technologies in medical field
resulting the rising in demand of life expectancy. In “Fall Detection System Using Accelerometer and Gyroscope Based on Smartphone” by Rakhman, Nugroho, Widyawan, and
Kurnianingsih (2014) proposed
a fall detection system using motion-based sensors using
smartphone. In the research, the proposed system used the accelerometer and
gyroscope sensors embedded in an android smartphone to send alert where it automatically
call the family members through an application. However, the system do not used
GPS technology in order to locate the whereabouts of user and the project also
did not implemented GSM technology to send the alert message. Thus, the user of
the smartphone needed to have internet data in order to receive the alert
in “Human Fall Detection Using
Three-Axis Accelerometer and ZigBee Technology”, Aquino, Magno, and Tuason (2012) developed a human fall detection and monitoring system based on
three-axis accelerometer and u using ZigBee technology. The proposed system is a
monitoring system which is an automatic device where after the fall is
detected, it will transmit data into a remote monitoring system automatically
without any intervention from the user. Nevertheless, the system cannot locate
the whereabouts of the fall victims since the project did not implemented GPS
technology. In addition, Aquino et al. (2012) stated that in the study, the developed device’s durability is
depending on the impact of falls.
the paper titled “A Real-time Fall Detection System for Maintenance Activities
in Indoor Environments” proposed by Triantafyllou et al. (2016) focused on a fall
detection in indoor environment which are real time, multi-space and multi-camera.
The process of fall demonstrated by using Hidden Markov Models (HMM) which based
on the falling bearing of a user such as the user’s velocity and area of variance.
According to the results of the experiments, there were only two false alarms
out of 5 similar falls occurred. However, this project did not used a GPS technology to track location when fall events occurred.
In addition, this project cannot be used in wider area range since it limits
for indoor environment.
the other hand, the proposed project by Jeon et al. (2017) titled “Self-Powered Fall Detection System Using Pressure Sensing
Triboelectric Nanogenerators” is a self-powered fall detection system based on
ambient using a triboelectric nanogenerators (TENG) array pressure sensor. Without
any external power source, each of the pressure sensing TENG array cell generated
analog signals when the contact of the two surfaces of cell is pressed. The
processor can categorize a falls or false falls events based on the analog
signals which can be done in a computer or a smartphone. This project also not
only can distinguish between falls and daily physical activities but can read
the output signal waveforms from different actions. The results showed 95.75%
classification of accuracy in identifying actual falls. But, this project can
only be applied in certain area, thus this project also did not applied GPS
technology which can track the location of falls victims.
the project proposed by Huynh, Nguyen, Tran, Nabili, and Tran
titled “Fall Detection System Using Combination of Accelerometer and Gyroscope”
used a wireless sensor system (WSS) based a combination of accelerometer sensor
and gyroscope sensor simultaneously to identify falls from normal Activities of
Daily Living (ADLs). This project compromised a tri-axis digital accelerometer
sensor, 3-axis digital gyroscope sensor, ARM microcontroller and a Wi-Fi module.
This project achieved an accuracy of 99.382% to distinguish between falling and
ADLs. However, this project did not implemented GPS technology to track
patient’s location and at the same time did not used GSM technology to receive
2.1 simplified the significance projects that relates with this proposed human
fall detection system project.
Table 2.2 Related Works
Fall Detection System
Using Accelerometer and Gyroscope Based
Arkham Zahri Rakhmani,
Lukito Edi Nugrohoi, Widyawani, Kurnianingsih
The proposed system
utilized a tri-axis accelerometer and
gyroscope contained on the smartphones. This study uses a
smartphone with android operating
system. While the sensors used are accelerometer and gyroscope sensor.
Did not implement GSM for sending short messages (SMS)
and did not used position determination using GPS
Human Fall Detection
Using Three-Axis Accelerometer and ZigBee Technology
Kimberli Anne M.
Aquino, John Lester S. Magno, Gizelle Ann C. Tuason
The proposed fall
detection device is a threshold-based
tri-axial accelerometer using MEMS
accelerometer technology. This device was embedded with a ZigBee wireless transmitter to
forward data to a remote system which manages the monitoring of the patient.
not have GPS
not have GSM
be used in wider area
A Real-time Fall Detection System
for Maintenance Activities in Indoor Environments
Triantafyllou, S. Krinidis, D. Ioannidis, I.N. Metaxa, C. Ziazios, D. Tzovaras)
focuses on the detection of fall incidents while it highlights the leverage
that such a system can provide to the human.
not used a GPS technology to track location
be used in wider area
Fall Detection System Using Pressure Sensing Triboelectric Nanogenerators.
Young-Hoon Nho, Sang-Jae Park, Weon-Guk Kim, Il-Woong Tcho, Daewon Kim,
Dong-Soo Kwon, Yang-Kyu Choi)
This paper proposed
an economical, ambient-based fall detection system based on a pressure
sensing triboelectric nanogenerator (TENG) array.
not use a portable and wearable device thus can only be used in a room
not used a GPS technology to track location
Fall Detection System
Using Combination of Accelerometer and Gyroscope
(Quoc T. Huynh, Uyen
D. Nguyen, Su V. Tran, Afshin Nabili, Binh Q. Tran)
paper studied on a wireless sensor system (WSS) based on accelerometer and
gyroscope fall detection system to collect data. The collection data program
is written in Matlab (Mathworks, Inc, Natick, MA). The program receives and
display real-time data from the WSS. The
WSS contains a set of ADXL345 (3-axis digital accelerometer sensor), ITG3200
(3-axis digital gyroscope sensor), MCU LPC17680
(ARM 32-bit cortex M3), and Wi-Fi module RN13
not implemented GPS technology to track patient’s location
not used GSM technology to receive alert
Development of Human
Fall Detection System Using Raspberry Pi with GPS and GSM Technology
paper focuses on a portable and wearable human fall detection system which is
inexpensive and easier to build. Once the falls event occurred, the device
will detect the falls and immediately sends an alert message along with the
victim’s location in a form of coordinates and link to Google Maps
Application on a smartphone.
be wear anywhere and anytime