You are on page 1of 1

Christian Llave TCOM151.

2 April 22, 2016

Reaction to Alcantara and Laguilless’s “Development of a Reliable Flood Sensor with an


SMS Alert System Using Both Contact and Non-contact Sensors”

Paul Alcantara and Lorenz Laguilles’s thesis aims to develop a hybrid flood sensor
consisting of a combination of the cotnact and non-contact sensors. The project is envisioned
to be utilized in urban areas. By interfacing a water sensor with an ultrasonic sensor, the
presence of water would be detected first before the ultrasonic sensor is activated to measure
the height of the water. For the data gathering, tests will be conducted to determine reliability
and performance of the sensors individually and collectively. Furthermore, the system to be
developed incorporates an SMS alert that will send a warning to users in real-time as the
flood level increases as measured by the ultrasonic sensor. Then, once the overall
functionality is developed, a field test for the system shall be conducted.
The study entailed having to construct a reliable flood sensor by interfacing contact
and non-contact sensors. They then needed to implement an SMS alert system indicating the
severity of the flood based on the distance measured by the flood sensor. They also conducted
a field test of the overall flood sensing system to be developed.
Their contact sensor made use of three metal probes with varying distances. I had
inquired Mr. Alcantara what determined their choice of the distance between the metal
probes. He said that the probe 13cm from the ground is based on their decision for the gutter
level. The succeeding probes are in accordance to the flood level warnings provided by the
Metropolitan Manila Development Authority.
I then asked if they considered using other sensors for their project. Mr. Alcantara
responded, saying they didn’t consider using any other sensor because they only needed a
contact and non-contact sensor. I asked for the reason. He stated that the sensor combination
was to detect false-positives. The contact sensor served as the control to the non-contact
sensor, which involved a PING sensor and a temperature sensor.
To better observe the accuracy of the non-contact sensor in detecting the rising of
water, measurements were done every 5cm of the ruler until 25 cm, and then 28 cm. The
rising of the water was momentarily stopped every time a measurement is done, hence, the
values should plateau at factors of 5.
The first trial made use of a calibrated value of 31cm. The difference in measurement
between the ruler and the ultrasonic sensor started with a deviation of 1cm at 15cm, 2cm at
20cm, 3cm at 25cm and 28cm. The second trial made use of 33cm as a calibrated value. The
difference was 1cm at 15cm and 20cm, 2cm at 25cm, and 3cm at 28cm. Their third trial
involved the use of a 27cm calibration. The difference was at 1cm at 15cm, 2cm at 20cm and
25cm, 3cm at 28cm.
The project revealed that as the surface of the water gets closer to the ultrasonic
sensor, the error increases linearly, depending on the calibration. These errors ought to be
compensated for.
Their project made use of an SMS alert system using an arduino GSM shield and a
SMART prepaid SIM. The counter value increases by 1 every time the 13cm flood height is
detected. After the counter value reaches 9, the SMS alert is sent. The counter value resets to
0 upon sending an SMS alert.
The project proved to be useful, with the need to compensate for the errors in the
measurement using the ultrasonic sensor.

You might also like