Flood Detection and Mitigition System

Using SMS Notifications

Muhamad Nabil Asyraf Ghazali1, Ansar Jamil1*, Jiwa Abdullah1,

Lukman Hanif Muhammad Audah1 and Rozlan Alias1
1
Faculty of Electrical and Electronic
Engineering,
Universiti Tun Hussein Onn Malaysia,
Batu Pahat, 86400, Johor,
MALAYSIA
[email protected]

1. Introduction
The main causes of flood are hydrological condition, metrological condition,
geographical condition, planning problem and environmental status due to human
activity etc. [1]. Hydrological condition arises due to the discharge of heavy rain fall
and the enormous water. The hydrological condition is directly and indirectly associated
with the metrological condition such as heavy rainfall, cyclone and storm. The
geographical condition is also a major factor for flooding condition, for example,
altitude of the place. Lower is the altitude of a place; the water flows from the
neighbouring higher altitude. The planning problem such as poor drainage, high
siltation in river, breaching of the embankments, spilling of floodwaters over them,
contribute to flooding situation. The environmental status due to human activity is a
significant cause for flooding situation, for example, deforestation, different type of
pollution and more buildings construction etc. [1]. Floods can have devastating
consequences and can have effects on the economy, environment, and people. During
floods (especially flash floods), roads, bridges, farms, houses and automobiles are
destroyed. The environment also suffers when floods occurred [2].
Today many flood detection systems in rural area in Malaysia used the conventional
way to predict the disaster; where the resident itself keeps watching water raises event
near the river side. This action is too risky because the area near the river become
dangerous for their safety, which can caused current water level data cannot be collected
causing a delay in notifying all resident [3]. The existing flood detecting system just
monitor water level for the main rivers but not providing an efficient alert system
especially to the resident near the river.

2. Related Work
Jaymala Patil and Anuja Kulkarni [4] came with an idea to enhance safety of track over
bridges against any unforeseen flash floods and breaches. The project used LPC 2148
Microcontroller as the controller, when it rise in water level will be sensed by the
electromechanical float sensor and sends the signal provided to the Central Processing
Unit (CPU). The project also included with GSM modem which turn to send Short
Message Service (SMS) notification.
Danny Hughes and Phil Greenwood from Computing Department, Infolab21,
Lancaster University [5] developed an intelligent and adaptable grid-based flood
monitoring and warning system. In their project, they used Gridstix platform in order
to describes a wireless sensor network (WSN) for flood warning which is not only
capable of integrating with remote fixed-network grids for computationally-intensive
flood modelling purposes, but it also capable of performing on-site flood modelling by
organising itself as a ‘local grid’.
 31

Nivethithaa.P and Karthiga S.R [6] from Krishnamy College of Engineering &
Technology, Cuddalore, India comes out with the role of satellite Zigbee technology in
flood monitoring and communication system. Zigbee is the latest wireless weather
monitoring technique. The vital role of the designed satellite Zigbee technology in flood
monitoring and communication sytem is based on mobile apps. The mobile apps is used
to communicate with the people during flood times and is to continuously monitor,
detect and report the environmental status to a control unit using water level sensor,
satellite Zigbee and the readings are displayed in it.
Anthone et al. [7] has described about an alternative network as a substitute to the
usual communication links which are unavailable during major disaster. They proposed
an alternative network for maintaining communications capabilities during major
natural disasters and other emergency situations by a system that utilizes Short Message
Service (SMS) of length up to 7bits over Wireless Mesh Sensor Networks (WMSNs).
This technique is relatively simple and inexpensive. Other related works on flood
detection and mitigation system are described in Table 1.

Table 1 - Existing flood detection and monitoring system

Project Description Pros and Cons
Early Flood Power supply AC to DC (transformer) Need AC power supply to turn on
Detection System Electrode
[8] 8051 series microcontroller
GSM module

Citizen Flood Small battery (lithium 3.6v) Small batteries

Detection Ultrasonic rangefinder Includes temperature sensor
Network [9] Ciseco SRF shield
deploys under bridge or overhang above
water
Automatic Flood 6 x 1.5 V battery Does not emit smoke, burning smell
Detection and Overhang above water or spark while being used
Warning Device, a Does not overheat
Prototype [10] The components of the device do not
blow up or short circuit.
The ALERT121A Integrated battery bracket that can Configurable as an alert transmitter
[11] accommodate up to 24 hour of and repeater
rechargeable battery Sealed, circular sensor connectors on
Style stand pipe installations lid of canister
Modular mounting of components for
easier field servicing and expansion.
Telemetry AC power supply Data updated 24 hour per day
System [12] Telemetry system Real time data accessed
Integrated data logging system tipping Radar sensors maintenance free
bucket rain gauge To ensure the sensor provide
Radar water level sensor accurate data by visually checking the
Live data sensor and measurement beam using
Bridge or overhead structure for securely a portable distance sensor or nearby
mounting the radar sensor staff gauge.
Fujitsu Flood Thermoelectric converter that Transmit water level info to the cloud
Detection [13] transforms energy from fluctuations of every 5mins
the manhole cover's heat into electricity Extends the time between battery
Thermoelectric converter replacement to five years from 10
Drainage pipes months
Smart Flood Bank of siren There were markings on the scale, but
Warning System Scale pole no indication as to which water level
[14] Near the river was the signal for precaution to be
taken
 32
Flood Protection Long life battery Early detection of moisture, water
Systems [15] Moisture leaks and flooding, which often go
Sensors are placed or installed in undetected for long periods.
potential leak, overflow, or flood-risk Protection against mold, which can be
locations throughout the home–this may a serious health risk.
be directly on the floor, in a cabinet Protection for elderly and children
beneath a sink or water heater, in the from falls caused by water leaks and
basement, or beside a bathtub. Once overflows.
water is detected, the sensor sets off the
alarm—likely a loud, audible sound
and/or blinking light
Flood Sensing Powered by solar panel Battery always rechargeable
Network by Arduino Uno R3 Only siren as alert warning
Arduino and WSN Xbee Inexpensive
[16] Near The River Solar panel thus benefits villagers,
because of a limited access to
electricity.

3. System Design
This system consists of ultrasonic sensor (HC-SR04) as the water level detection. The
Arduino UNO is selected as the system microcontroller and connected to the GSM
modem. Both of the Arduino UNO and GSM modem supplied by a battery. The system
intended are to provide information of the current water level status in the selected area
and give an alert notification to the user if the water level rises that would cause a flood.
This system is user friendly and be attained anywhere as long as mobile communication
coverage is available. Figure 1 shows the block diagram of the system.

Fig. 1 - Block diagram of the system with GSM/GPRS module

 33

Figure 2 shows deployment of the system to detect current water level of a river,
drain, or etc. Table 2 indicate the parameters that being used.

Fig. 2 - Visualization of device implementation

Table 2 - Parameters for water level calculation

Parameter Description Notes

A, B Distance from cliff (1 meter) -
Cdeploy Distance of sensor to water surface during -
deployment
Ccurrent Current distance of sensor to water surface
Ddeploy Level of water during deployment -
E STANDBY level 40 cm below the bank
(Green Indicator/
F EVACUATE level 20 cm below the bank
(Yellow Indicator)
G Danger Level Above 20 cm from from the
(Red Indicator) bank

In order to deploy the system, a procedure should be followed as the following:

1. Measure the current water depth (Ddeploy) using a benchmark stick.

2. The device must be placed 1 meter upper and 1 meter apart from the water bank
(refer A and B).
3. When placing the device, make sure that the ultrasonic sensor of the device is
facing the surface of the water.
4. Set the depth of water (Ddeploy) and current reading of ultrasonic sensor (Cdeploy)
during deployment manually in the source code and program the device.
5. The ultrasonic sensor will detect the rising of water level. The current water
level (Dcurrent) is calculated using the equation (1).

Dcurrent = Ddeploy + (Cdeploy – Ccurrent) (1)

 34

6. The current water level reading (Dcurrent) is send to the user using SMS.
7. Alert notifications are sent to the user when the water level reaches STANDBY,
EVACUATE and DANGER level.

4. Experimental Works
4.1 Indoor Experiment

Figure 3 shows the indoor measurement setup that consist of a measuring stick, a big
container, a laptop and a table. The important parameters about the indoor experiment
setup is shown in Table 3. The purpose of these experiments are to study about the
accuracy of the system, the effect of wave and floating materials on the water level
reading of the system.

Fig. 3 - Tools that have been set up before doing the experiment

Table 3 - Parameters for indoor experiment setup

Parameter Values

Distance between sensor and water surface (C) 68 cm

Current Depth (D) 2 cm

Box Container Depth 29 cm

35

4.2 Actual Implementation

The purpose of these experiment is to study the actual performance of the system. Two
experiments have been conducted in selected areas within UTHM. Figure 4 and Figure
5, shows the actual experiments were conducted at two different locations in UTHM
campus: drain between block E14 & E15 and drain in front of Diploma Study center
(PPD). In this experiment, the system collects level of the water for three times each
day. Each time, ultrasonic sensor collects data for three times every 10 minutes Average
water level is calculated based on these readings. The water level status is send along
with the sensor reading via SMS notification, as shown in Figure 6.

Fig. 4 - Actual experiment setup at the drain between block E14 & E15

Fig. 5 - Actual experiment setup at the drain in front of Diploma Study center (PPD)
 36

Fig. 6 - Water level readings sent via SMS

5. Data Analysis for Indoor Experimental

5.1 Accuracy of the Water Level Reading

The purpose of this experiment is to determine the accuracy of the water level reading
collected by the system. As comparison, the actual water level is measured using
measuring stick. Initially, the water container is filled with water until it reach to the
level equal to 2cm. The system collects about 14 readings of water level. Then, each
sensor reading is sent to a laptop using serial communication. Then, t water level is
increased for each 2 cm. The same step is repeated for each increase of the water level.
Figure 7 shows the water level reading from the system compared to the actual
measurement. Based on the result, the water level reading from the same is almost
consistent with the actual measurement and the highest standard error deviation is
0.33.
Accuracy of Water Level Reading

30
25
Water level reading (cm)

20
15
Actual
10
5 Average Sensor
Reading
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Number of readings

Fig. 7 - Graph of water level reading through measuring stick

against serial monitor
37

5.2 Effect of Wave on the Water Level Reading

This experiment is conducted to understand the effect of wave of the water level
reading. Three different levels of water are tested: 9cm, 17cm and 25 cm. In this
experiment, wave is generated in this experiment by shaking the water using an empty
bottle. Small part of the bottle is sunk into the water. The bottle is rotated for about 20
times per minute to create wave. The water level reading is taken for a duration of 150
seconds. The time interval for each reading is 5 seconds. The experiment is repeated for
5 times. Figure 8 shows the plotted graph of water level reading over time for each
water level.

Effect of Wave on the Water Level Reading

30

25
Water level reading (cm)

20

15

10

0
0 25 50 75 100 125 150
Time (s)
Fig. 8 - Graph effect of slow wave on standby water level reading

Based on Figure 8, the result shows that wave has a significant effect on the water
level reading. The water level data is fluctuated. It is because wave on the surface of
the water may increase or reduce the water level reading. The result also shows that, the
higher the water level, effect of wave of the water level become more significant.

5.3 Effect of Floating Materials on the Water Level Reading

This purpose of the experiment is to determine effect of floating materials on the water
level reading of the system. Four floating materials are used: bottle, polystyrene, biscuit
container and tree bark as shown in Table 4. The level of water is fixed at value of 23
cm.
 38

Table 4 - Floating materials

Materials Plastic bottle Polystyrene Biscuits Tree Bark
Container
Thickness 9 cm 6 cm 14 cm 5 cm

Picture

Effect of Floating Materials on Water Level Reading

40
Water level reading (cm)

30

20

10

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29
No. of reading
No Material Bottle Polystyrene
Biscuit Tin Tree Bark
Fig. 9 - Graph of the effect of floating materials on water level reading

Figure 9 shows water level readings over time for all the floating materials. Overall,
the results shows that the water level reading of the system affected by the floating
materials. It is found that the floating materials make the water level readings becomes
inaccurate. This is because, the floating materials exist between the ultrasound sensor
and the surface of water. This will caused the emitted signal of the ultrasound sensor is
reflected by the floating materials and not the surface of the water that gives an accurate
reading. The error in water level reading depends on the thickness of the floating
materials. The higher the thickness of the floating material, the larger the error of the
water level readings. In addition of that, if the floating materials have uneven or bumpy
surface, can caused an error in the reading because of the signal is not reflected back to
the ultrasound sensor. As an example, water level reading for tree bark is in error
because the reading is below the actual water level.
39

6. Data Analysis of Actual Implementation in UTHM

6.1 Experiment Result for Drain between Block E4 & E15, UTHM

Figure 10 shows water level reading for drain between block E4 & E15, UTHM. The
experiment was conducted for 8 days from 23/2/2017 until 30/11/207. Based on the
Figure 10, the highest water level was recorded on 29/11/2017 in the afternoon, which
is equal to 41.67cm due to heavy rain. The lowest water level was on 25/11/2017, which
the water level reading equal to 23.63cm height.

Fig. 10 - Graph of water level reading at the drain E14 & E15

6.2 Experiment Result for the Drain in front of PPD, UTHM

Figure 11 shows the water level reading for the drain in front of PPD. The experiment
was conducted for seven days from 29/11/2019 until 5/12/2017. It is found that the
water level decreased throughout the week due to no rain. The highest water level was
recorded in the afternoon on 29/11/2017, which is equal to 43.67cm. The lowest water
level reading was recorded in the afternoon on 5/12/17 with the reading value equal to
34cm.
 40

Fig. 11 - The graph of water level reading at the drain in front of PPD

7. Conclusion
As a conclusion, the flood detection and mitigation system has been developed to
monitor water level of river, drain, stretch or canal, and provide an early warning of
flood. The system is built using Arduino Uno microcontroller. Ultrasound sensor (HC-
SR04) is connected to Arduino to detect current water level. GSM/GPRS SIM900
modem is included in the design in order to allow water level reading and notification
message to be delivered via SMS. In order to determine the performance of the system,
few experimental works had been conducted. These experimental work includes to
determine the accuracy of water level reading, the effect wave and floating materials on
the water level reading. The actual experimental are conducted in UTHM area. As a
recommendation, the system should be improved to provide more accurate water level
readings in the condition of wave and floating material.
41

References

1. I. Priyadarshinee, “A Model for Flood Prediction and Prevention using Wireless Sensor
Network,” International Conference on Emergent Trends in Computing and Communication
(ETCC 2015), pp. 22–30, 2015.
2. K. Prasanti, C. J. Lakshmi, and C. S. Savalam, “A prototype controlling of barrage gates
using arduino under flood conditions,” pp. 579–583, 2016.
3. A. Engineering, B. Vidya, K. Poonam, G. Priyanka, D. Gaurav, and P. A. S. Chandgude,
“Water Level Monitoring System In Real Time Mode Using WSN,” Int. J. Emerg. Technol.
Adv. Eng., vol. 6, no. 9, pp. 212–214, 2016.
4. Patil, Jaymala, and Anuja Kulkarni. "Wireless Sensor Network Using Flood Monitoring."
International Journal of Computer Science and Mobile Computing, IJCSMC 2.11 (2013):
297-302.
5. Hughes, Danny, et al. "Gridstix: Supporting flood prediction using embedded hardware and
next generation grid middleware." World of Wireless, Mobile and Multimedia Networks,
2006. WoWMoM 2006. International Symposium on a. IEEE, 2006.
6. Nivethitha, P., S. R. Karthiga, and C. Reikha. "The Role of Satellite Zigbee Technology in
Flood Monitoring and Communication System." 2017.
7. Anthone Andrade, Mynor Vinicio, and Satoru Oishi. "A Wireless Mesh Sensor Network
Framework For River Flood Detection And Emergency Communications In Case Of
Disaster." 2014.
8. Permut, Alan R., Albert A. Permut, and Ronald M. Permut. "Early flood warning system."
U.S. Patent No. 4,153,881. 8 May 1979.
9. “Flood Network by PhotoScape” [Online]. Available: http://www.postscapes.com/citizen-
flood-detection-network/ [Accessed: 24-Feb-2017]
10. “Automatic Flood Detection and Warning Device a Prototype: [Online].
Available:https://www.slideshare.net/shainamavreenvillaroza/automatic-flood-detection-
and-warning-device-a-prototype [Accessed: 24-Feb-2017]
11. “Alert Flood Warning System” [Online].
Available: https://www.campbellsci.com.au/alert121a [Accessed: 25-Feb-2017]
12. “Flood Warning System” [Online].
Available:http://www.fondriest.com/environmentalmeasurements/environmental-
monitoring-applications/flood-warning-systems/ [Accessed: 25-Feb-2017]
13. “Fujitsu Flood Detection Device” [Online].
Available: http://www.fujitsu.com/global/about/resources/news/pressreleases/2016/0815-
01.html [Accessed: 25-Feb-2017]
14. “Smart Flood Warning” [Online]. Available: https://www.campbellsci.com.au/news-17-q4-
smart-flood-warning [Accessed: 25-Feb-2017]
15. “Flood Protection System” [Online]. Available: https://www.safety.com/p/flood-protection-
systems/ [Accessed: 25-Feb-2017]
16. Vunabandi, Vianney, et al. "Flood Sensing Framework by Arduino and Wireless Sensor
Network in Rural-Rwanda." Software Engineering, Artificial Intelligence, Networking and
Parallel/Distributed Computing (SNPD), 2015 16th IEEE/ACIS International Conference
on. IEEE, 2015.