REVOLUTIONIZING AQUACULTURE: ENHANCING

SUSTAINABILITY THROUGH IOT AND SUPERVISED MACHINE

LEARNING IN RAS

STUDENT NAME: Minerva M. Fiesta

STUDENT NUMBER: 21-3944-222
COURSE NAME: Doctor in Information technology
COLLEGE: College of Information Technology and Computer Science
COURSE CODE: DIT D1 – Dissertation 1

PROFESSOR: THELMA D. PALAOAG

DATE OF SUBMISSION: 29 07 2023

CONTENTS
ABSTRACT........................................................................................................................................................................3

INTRODUCTION..............................................................................................................................................................4

PROBLEM STATEMENT................................................................................................................................................5

OVERVIEW........................................................................................................................................................................5
RESEARCH QUESTION.......................................................................................................................................................6

OBJECTIVES AND AIMS................................................................................................................................................6

OVERALL OBJECTIVE........................................................................................................................................................6
SPECIFIC AIMS..................................................................................................................................................................6

BACKGROUND AND SIGNIFICANCE.........................................................................................................................6

RESEARCH DESIGN AND METHODS........................................................................................................................7

OVERVIEW........................................................................................................................................................................7
POPULATION AND STUDY SAMPLE...................................................................................................................................7
SAMPLE SIZE AND SELECTION OF SAMPLE......................................................................................................................8
SOURCES OF DATA...........................................................................................................................................................8
COLLECTION OF DATA......................................................................................................................................................8
DATA ANALYSIS STRATEGIES..........................................................................................................................................8
TIMEFRAMES.....................................................................................................................................................................8

STRENGTHS AND WEAKNESSES OF THE STUDY.................................................................................................9

BUDGET AND MOTIVATION.......................................................................................................................................9

REFERENCES.................................................................................................................................................................10

2
 ABSTRACT
Aquaculture is vital for meeting the increasing global demand for seafood, but it faces challenges
such as environmental impact and inefficient resource management. Recirculating Aquaculture
Systems (RAS) offer a sustainable solution by creating a controlled environment for fish farming.
However, optimizing RAS operations requires continuous monitoring and management of various
factors. The Internet of Things (IoT) and supervised machine learning can revolutionize
aquaculture by providing real-time data analysis and predictive modeling. This dissertation aims to
investigate the factors influencing sustainable freshwater aquaculture using supervised machine
learning in RAS. A comprehensive IoT-based monitoring system is designed and implemented to
collect real-time data on environmental parameters. Supervised machine learning algorithms are
then applied to analyze the data, identify influential factors, and build predictive models. The
developed model’s water quality control in aquaculture operations. Rigorous validation and testing
ensure the accuracy and effectiveness of the models. The results contribute to sustainable
aquaculture practices by guiding decisions on resource management, productivity improvement,
and environmental impact reduction. The integration of IoT and supervised machine learning has
the potential to revolutionize the aquaculture industry, paving the way for a more efficient and
sustainable future.

3
 INTRODUCTION
A country's economic development largely relies on agricultural products, as they serve as
the primary source of food and essential raw materials. To accomplish the world's development
objectives, establishing healthy, sustainable, and inclusive food systems is of utmost importance.
Agriculture significantly influences global trade as it is interconnected with various sectors of the
economy, leading to job creation and fostering economic development. This is due to the
innovative use of technology and advanced farm management practices by producers in these
nations, resulting in increased agricultural productivity and profitability (The World bank, 2023).
Aquaculture can be seen as a close aquatic counterpart to agriculture, involving the
cultivation of specific marine and freshwater organisms to supplement natural resources. (Andres
R.F.T. von Brandt, 2023). The Recirculating Aquaculture System (RAS) is an advanced method of
fish farming that enables high-density fish production while being environmentally sustainable. In
addition, Recirculating Aquaculture Systems (RAS) is considered as one of the options for
adapting the impact of climate change on fish production while ensuring environmental
sustainability. Unlike conventional aquaculture, Recirculating Aquaculture Systems (RAS) do not
present ecological risks such as the decline in biodiversity caused by confined fish or the
transmission of viruses and parasites. Instead, RAS is an environmentally beneficial, water-
efficient, and highly productive form of intensive agriculture. (Ahmed & Turchini, 2021).
The primary focus of this dissertation proposal revolves around two essential components:
the Internet of Things (IoT) and supervised machine learning. IoT, a technology that has emerged
during the digital era, finds diverse applications across various domains. (Pitakphongmetha et al.,
2021). By offering real-time data, these interconnected devices empower aquaculturists to make
informed choices and enhance the production process. The integration of advanced information
technologies like the Internet of Things, cloud computing, big data, and AI has resulted in notable
advancements in aquaculture and paved the path for intelligent fishing production.(Bradley et al.,
2019).
The last five years have seen the adoption of machine learning techniques and algorithms
in intelligent fish farming, and studies of the results have been done. (Zhao et al., 2021).
Supervised machine learning algorithms can be employed to evaluate data collected by IoT
devices and offer predictions or suggestions to enhance the efficiency and sustainability of the
system. This may involve enhancing feeding schedules, early disease detection, or more effective
water quality management. Machine learning technology has become increasingly prevalent in
aquaculture, providing new opportunities for digital fish farming in light of advancements in
automation and intelligence.
As an agricultural nation, the Philippines must prioritize investments in developing resilient
and sustainable agriculture and food systems capable of withstanding disasters. By doing so, the

4
country can attain food self-sufficiency, promote the growth of rural communities, and elevate
farmers' income. Agriculture holds a significant role in the economy, contributing approximately
400% to the GDP and accounting for two-thirds of all job opportunities. (Agri Farming, 2022).
Improving farm productivity is essential in order to increase farm profitability and to provide the
rapidly growing demand of food caused by rapid population growth all over the world. According to
estimates compiled by the Food and Agriculture Organization (FAO), by 2050 we will need to
produce 60 per cent more food to feed a world population of 9.3 billion (Johannesson, 2019). In
conclusion, recirculating aquaculture systems offer several benefits over traditional aquaculture
methods, including reduced environmental impact, improved water quality, and increased
efficiency (Aqua Farm, 2023).
Hence, there is a need to revolutionize Aquaculture by Enhancing Sustainability through
Internet of Things and Supervised Machine Learning in a Recirculating Aquaculture System.

PROBLEM STATEMENT

Overview
Fish farming, also known as aquaculture, has emerged as a critical industry in meeting the
world's expanding need for protein food. With aquaculture currently accounting for more than half
of global fish consumption, this industry plays a critical role in ensuring food security and
supporting livelihoods for millions of people. Climate change, on the other hand, has emerged as a
key danger to fish farming, offering many difficulties to the global sustainability and productivity of
aquaculture operations.
The freshwater aquaculture industry faces challenges in maintaining optimal water quality
conditions to ensure the growth and survival of aquaculture products due to the effect brought by
climate change. Traditional monitoring techniques frequently have a limited capacity to deliver
complete and real-time data on water parameters. Fish growers therefore strive to maintain the
ideal circumstances required for the accomplishment of their aquaculture operations and year-
round fish output. The lack of comprehensive understanding of how water parameters affect the
growth rate and survivability of aquaculture products further hampers the implementation of
effective strategies for small-scale fish farming, particularly in upland areas.
Hence, the primary objective of this study is to evaluate and ensure the ideal water quality
conditions in Recirculating Aquaculture Systems (RAS), thereby assisting fish farmers in
maintaining favourable environments for their aquaculture products. Through continuous
monitoring and analysis of water parameters, the study aims to improve the growth rate and
survival of aquaculture products, with a specific focus on small-scale fish farming in upland areas.
Furthermore, employing suitable machine learning algorithms for the analysis of water parameter

5
data in RAS will then provide valuable insights and recommendations to support the improvement
of aquaculture practices and promote the success of fish farming ventures in these regions.

Research Question
1. What are the design considerations and technological requirements for developing a device
that effectively acquires accurate and reliable water parameters data in a RAS?
2. How can a machine learning algorithm be utilized to develop a water quality evaluation
model for assessing the aquaculture survivability in Recirculating Aquaculture Systems
(RAS)?
3. What is the extent of usability of the developed system?

OBJECTIVES AND AIMS

Overall Objective

The general objective of this study is to Develop a device to monitor water parameters in a
recirculating aquaculture system (RAS). and analyze the data using suitable machine learning
algorithms.

Specific Aims
1. To develop a device that acquires water parameters data.
2. To develop a water quality evaluation model using a machine learning algorithm that
effectively evaluates the survivability of aquaculture in Recirculating Aquaculture Systems
(RAS).
3. To evaluate the extent of usability of the developed system.

BACKGROUND AND SIGNIFICANCE

The remarkable expansion of aquaculture has led to a historic peak in global fisheries and
aquaculture output, with aquatic foods playing a progressively vital role in ensuring food security
and nutrition in the 21st century (FAO, 2022). Aquaculture serves as the primary fish source in the
country, The most prevalent species cultivated in the country are milkfish, tilapia (predominantly
Nile tilapia, Oreochromis niloticus), and shrimp. (National Fisheries Development Board, 2023).
However, the sustainability and efficiency of freshwater aquaculture operations face significant
challenges. Traditional aquaculture practices often lack real-time monitoring capabilities, leading to
suboptimal resource utilization, increased environmental risks, and reduced productivity.
Additionally, the complex relationship of various factors, such as water quality, fish behavior, and
system performance, makes it difficult to optimize aquaculture outcomes.

6
 Recirculating aquaculture systems (RAS) that incorporate Internet of Things (IoT) and
supervised machine learning offer a revolutionary means of overcoming these difficulties. Sensors
and data analytics offered by IoT enable complete, real-time monitoring of critical parameters,
allowing for faster response and improved resource management. Aquaculture systems can be
better understood and managed by employing supervised machine learning algorithms, which can
examine huge datasets, determine patterns, and develop predictive models. This study focuses on
expanding seafood demand, the environmental consequences of conventional fishing practices,
and the limitations of current aquaculture approaches. This study is significant because it has the
potential to transform aquaculture techniques, promote sustainability, and significantly improve the
efficiency and productivity of Recirculating Aquaculture Systems (RAS).
This research employs Internet of Things sensors to enable real-time monitoring of critical
variables such as water quality, temperature, and feeding systems. Data analysis, predictive
insights, and data-driven decision-making are all possible with the application of supervised
machine learning algorithms in aquaculture operations. This integration has the potential to
improve resource utilization, lower environmental impact, reduce disease outbreaks, and boost
overall production efficiency. The study's importance reaches various stakeholders. Aquaculturists
and farmers can benefit from improved management methods, better productivity, and cost
savings. Environmental sustainability is enhanced by more effective resource utilization and
reduced environmental impact. Furthermore, consumers have access to sustainably obtained
seafood, which helps to meet expanding demand while also protecting the well-being of aquatic
animals. Overall, this study highlights the need for innovative approaches to revolutionize
aquaculture practices and enhance sustainability. By harnessing the potential of IoT and
supervised machine learning in RAS, this research contributes to the development of practical
solutions for a more efficient, environmentally friendly, and sustainable aquaculture industry.

RESEARCH DESIGN AND METHODS

Overview
The study will employ a descriptive-developmental study design method. This design will be
used to collect data on existing situations. The main goal is to collect and interpret information on
the nature of the situation as it exists at the time of the study.
The study is intended to be developmental. The proposed system will be designed to evaluate
water quality in a Recirculated Aquaculture System (RAS), thereby promoting a healthy
environment for aquacultures, lowering aquaculture mortality during cultivation, and achieving
maximum growth by maintaining RAS water quality based on the aquacultured optimum water
quality preference.

7
Population and Study Sample
This study will use the super-intensive tilapia culture system using 32 to 46 days old tilapia
(starter) with approximately 50 grams weight.

Sample Size and Selection of Sample

Sampling plays a crucial role in data collection, enabling predictions through statistical
inference. In this study, the researcher will employ purposive sampling, a type of non-probability
sampling, which involves selecting subjects based on the researcher's judgment regarding their
suitability and alignment with the study's criteria. The researcher will identify people who will take
part in the evaluation. It is important to identify the respondents who will evaluate the system
because they can easily cope with the researcher’s objectives. With this sampling technique, the
researcher will use a survey questionnaire to be answered by the chosen respondents. This
questionnaire will determine the overall functionality, usability, reliability, and performance of the
system.

Sources of Data
Through the help of the local government office Bureau of Fisheries and Aquatic Resources
BFAR and DA, the researcher will be able to gain knowledge and learn more about aquaculture,
especially about Tilapia which is the chosen aquaculture for this study. Interviews and site visits at
training centers of The Bureau of Fisheries and Aquatic Resources (BFAR) will be conducted. Data
will also be obtained from previous studies, journals and researches on Recirculating aquaculture
system using IoT and Supervised machine learning.

Collection of Data
Manual Collection of Data – using measuring devices
Automated Collection of data – using sensors.
Data in this study will be obtained through a series of in-depth interviews with the fish
farmers and personnels from BFAR in Alfonso Lista. The study will also use observation and
experimentation, a review of documents on the existing practices and procedures, the review of
previous studies, journals and research on recirculating aquaculture systems using IoT and
supervised machine learning.

Data Analysis Strategies

This study will consider three machine-learning algorithms to evaluate water quality in the
RAS and these include Decision tree, Fuzzy logic, and Linear regression. These algorithms will be

8
compared and contrast based on the algorithm’s purpose, advantages, disadvantages, and
implementation to find the most suitable algorithm to be used in this study.

Timeframes
The time of this study would be from May 2023 to February 2024
SY 2023-2024
Activity
MAY JUNE JULY AUG SEPT OCT NOV DEC JAN FEB MAR APR
Proposal Writing
Write Chapter 1- Introduction/RRL
Data Collection – Interview/Observation
Data analysis and Interpretation
Write Chapter 2 - Methodology
Project Development
Submit paper for publication
Write Chapter 3 – Discussion of Findings
Submit paper for publication
Project finalization/implementation
Write Chapter 4 – Conclusion and Recommendation

STRENGTHS AND WEAKNESSES OF THE STUDY

The study investigates the application of IoT and supervised machine learning in freshwater
aquaculture as a new strategy to improve aquaculture technologies' sustainability and operational
efficiency. The study makes use of Internet of Things (IoT) technology to accomplish real-time
monitoring of important aquaculture system characteristics, allowing for timely interventions and
data-driven decision-making to improve resource management and increase production efficiency.
The study applies supervised machine learning algorithms to examine massive datasets, identify
patterns, and develop prediction models. This technique enables a more in-depth understanding of
the numerous factors driving aquaculture performance and promotes well-informed decision-
making. The study addresses real-world difficulties faced by the freshwater aquaculture business,
such as poor resource utilization, environmental concerns, and dropping production. The study's
goal is to provide useful insights and recommendations to improve aquaculture operations and
support industry success.
The study's findings are dependent on the quantity and quality of data available for
analysis. A lack of comprehensive and credible datasets may restrict the accuracy and significance
of the findings. Implementing IoT and machine learning technologies in aquaculture systems may
require significant infrastructure and technical expertise. Limited resources and technological
constraints could hinder the widespread adoption of the proposed solutions.

BUDGET AND MOTIVATION

Expenditure Description Budget Request Justification of Expenditure
A. Supplies and Services
Consumable supplies Php 3,000.00 Bond paper, pen/pencil, printing, photocopying
9
 and others
Communication expenses Php 3,000.00 For gcash mobile call/text loads to contact
project respondents/ participants.
Travel expenses Php 5,000.00
B. Equipment
Hardware devices and Php 70,000.00 Project/system development
peripherals
TOTAL Php 81,000.00

Overall, A sustainable aquaculture using innovative technology to increase productivity is

needed to provide the rapidly growing demand of food caused by rapid population growth all over
the world. The research is driven by the goal of transforming aquaculture practices, enhancing
sustainability, and providing data-driven solutions to the challenges faced by the aquaculture
industry. By exploring the potential of IoT and supervised machine learning in RAS, the research
seeks to pave the way for a more efficient, environmentally conscious, and economically viable
future for aquaculture.

REFERENCES
Agri Farming. (n.d.). How to Improve Agriculture in the Philippines, Ways, Ideas, and Tips.
https://www.agrifarming.in/how-to-improve-agriculture-in-the-philippines-ways-ideas-and-tips
Ahmed, N., & Turchini, G. M. (2021). Recirculating aquaculture systems (RAS): Environmental
solution and climate change adaptation. Journal of Cleaner Production, 297, 126604.
https://doi.org/10.1016/j.jclepro.2021.126604
Andres R.F.T. von Brandt, C. H. A. (2023). https://www.britannica.com/topic/aquaculture. In
Agriculture & Agricultural Technology (pp. 1–18).
Aqua Farm. (2023). Revolutionizing Aquaculture : The Benefits and Functionality of Recirculating
Aquaculture Systems. 1–5.
Bradley, D., Merrifield, M., Miller, K. M., Lomonico, S., Wilson, J. R., & Gleason, M. G. (2019).
Opportunities to improve fisheries management through innovative technology and advanced
data systems. Fish and Fisheries, 20(3), 564–583. https://doi.org/10.1111/faf.12361
FAO. (2022). The State of World Fisheries and Aquaculture 2022. The State of World Fisheries
and Aquaculture 2022, 4–9. https://doi.org/10.4060/cc0461en
Johannesson, T. (2019). (12) By 2030, we will need 50% more food, 45% more energy, and 30%
more water | LinkedIn. 1–7. https://www.linkedin.com/pulse/2030-we-need-50-more-food-45-
energy-30-water-torfi-johannesson/
National Fisheries Development Board. (2023). Recent Trends in Aquaculture Recirculatory
Aquaculture System (Ras). Ministry of Fisheries, Animal Husbandry & Dairying, 040.
Pitakphongmetha, J., Suntiamorntut, W., & Charoenpanyasak, S. (2021). Internet of things for
aquaculture in smart crab farming. Journal of Physics: Conference Series, 1834(1).
https://doi.org/10.1088/1742-6596/1834/1/012005
10
Zhao, S., Zhang, S., Liu, J., Wang, H., Zhu, J., Li, D., & Zhao, R. (2021). Application of machine
learning in intelligent fish aquaculture: A review. Aquaculture, 540, 736724.
https://doi.org/https://doi.org/10.1016/j.aquaculture.2021.736724

11