11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Design and Optimization of a Bio-Composter System
Using Genetic Algorithm
1

5

Joshua Vivern Decano, 2Channon Jude Dominguez, 3Trisia Ann Luna, 4 Abraham Macayan Jr.
Lorcelie B. Taclan, 6Melquiades B. Garrino, 7Elmer P. Joaquin, 8Edwin A. Balila, 9Melvin T. Valdez
,10Jonalyn D. Castano*

Adventist University of the Philippines
jdcastano@aup.edu.ph

ABSTRACT
The Philippine agricultural sector's reliance on costly imported chemical fertilizers creates
economic burdens and compromises long-term sustainability. To address this, organic
composting provides a local, eco-friendly alternative, but its success relies on achieving
consistent, high-quality compost. This study aimed to design and construct a small-scale
composter system that utilizes Genetic Algorithm in the quality optimization of composter
operational parameters and integrate monitoring and remote capabilities. The study
specifically aimed to develop an optimization model using a genetic algorithm to a) to
determine the best composter operational parameters for each compost material (soy pulp,
vegetable peelings, coffee grounds) and b) to maximize composting efficiency, minimize
resource usage and enhance overall sustainability of the system. The resulting NPK values of
each compost material were compared to the commercialized fertilizer. The genetic algorithm
achieved realistic results in determining the best composter operational parameters for each
compost material with the given target NPK(Nitrogen, Phosphorus, Potassium) values.
Findings demonstrated that increased runtime significantly improves compost quality, while
optimal rotation speed yields diminishing returns beyond a certain threshold. Composter
system software also provided ease of use through an interface for composter control and
monitoring systems. To further improve the system, integration of advanced automation and
upgrading the motor power capabilities are recommended.
Keywords:

Genetic Algorithm, Composter, Fertilizer, NPK, Optimization

INTRODUCTION
The agricultural sector's reliance on chemical fertilizers to boost crop yields is evident
in the Philippines, with 2018 sales hitting 1.2 million metric tons(Authority, 2020). However,
local farmers often underutilize nitrogen due to high fertilizer costs, caused by the decline in
domestic production(David & Inocencio, 2019). This cost barrier, combined with limited
irrigation access, reduces fertilizer efficiency and crop yields. To address these issues, the
Philippine government promotes organic composting via the Composting Facilities for
Biodegradable Waste (CFBW) program. However, the quality monitoring of compost remains
a challenge.
1473

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

The prohibitive cost of fertilizers has been identified as a significant barrier, leading to
reduced application rates that fall below the levels necessary for optimal crop production
(Crisostomo, 2018). Additionally, limited access to irrigation further exacerbates the challenge,
restricting the effective use of fertilizers (Agriculture, 2020). These factors contribute to
inefficiencies in fertilizer management, which, if addressed, could potentially increase rice
yields significantly ((IRRI), 2021).
In response to these challenges, the Philippine government, through the Bureau of Soils
and Water Management (BSWM), has promoted the use of organic composting as an
alternative to chemical fertilizers. Initiatives such as the distribution of Rotary Composters and
Shredder Machines under the Composting Facilities for Biodegradable Waste (CFBW)
program aim to enhance organic waste management ((BSWM), 2021). However, these
initiatives often lack advanced systems for monitoring the quality of compost produced, which
is crucial for ensuring the effectiveness of compost as a fertilizer alternative.

LITERATURE REVIEW
Recent discussions in environmental science underscore the critical balance
required in nitrogen fertilization, highlighting the detrimental effects of overfertilization on
both soil health and the environment. Grell (2021) advocates for the development of multisensor integrated bio-waste composter systems that utilize machine learning to assess compost
quality against commercial fertilizers, pointing out the need for precise nutrient management.
Kim et al. (2018) further emphasizes the role of IoT-enabled remote sensors in
monitoring soil nutrients, such as nitrogen. These technologies enable dynamic adjustments
based on continuous data streams, which are crucial for maintaining optimal soil health (Grell,
et al., 2021).
Focusing on the technical advancements in composting, recent studies have integrated
genetic algorithms to enhance composting outcomes. For instance, a significant study by
Guochao Ding et al. (2023) utilized a CGA-BP neural network-based model to optimize the
aeration oxygen supply in aerobic composting. This model significantly improved the
compost's maturity by optimizing oxygen levels, demonstrating the efficacy of combining
genetic algorithms with neural networks to refine composting processes (Guochao Ding, 2023).
In another innovative approach, a study involved the use of genetic algorithms to
optimize humic acid content in GLR composting (Chun-Fang Shi, 2022). By adjusting various
composting parameters, this study highlighted the potential of genetic algorithms in enhancing
specific nutrient profiles within composts, underscoring their versatility in environmental
management applications.
This study focuses on developing a biowaste composter system using genetic algorithms and
sensor technology to improve compost quality. Recent advancements, such as multi-sensor
integration and IoT-enabled remote sensors, are being employed to monitor soil nutrients and
1474

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

optimize compost production. Integrating machine learning and genetic algorithms has shown
promise in improving composting processes, nutrient profiles, and sustainability in agricultural
practices.

METHODS
Materials: Devices, Technologies, and Equipment
1. Arduino Mega 2560 – The Arduino Mega 2560 is a microcontroller board based on the
ATmega2560 chip.Its primary role within our system involves managing electronic
components, interfacing with sensors for additional parameter monitoring, and executing
programmed tasks essential for optimizing the composting process using genetic
algorithms.
2. Raspberry Pi 4 Model B – A single-board computer designed for diverse applications,
this device supports program execution and data processing which is crucial for
implementing genetic algorithms in bio-composting.
3. Stepper Motor Driver TB6600 – is a driver specifically designed to control stepper
motors. A stepper motor is a type of electric motor that rotates in discrete steps. The
TB6600 driver serves as a vital link between our system's microcontroller, such as an
Arduino, and the stepper motor, enabling accurate and controlled movement essential for
optimizing composting processes.
4. Stepper Motor Nema 23 30kg-cm – is a standardized stepper motor with a flange
diameter of around 2.3 inches and a holding torque of 30 kilogram-centimeters. This
specific motor is selected for its ability to provide precise positioning and sufficient
torque output, crucial for the operational requirements of the bio-composter system.
5. Temperature and Humidity Sensor – The sensor employed in our bio-composter
system offers accurate temperature and humidity measurements with rapid response
times, high reliability, and resistance to interference.
6. Hall Effect Sensor – Integrated circuits that transduce magnetic fields to electrical
signals with accuracy, consistency, and reliability. It is a feature for the rotating
composter to precisely align the docking system, to ensure convenient and seamless
extraction and insertion of compost.
7. Rotary Composter bin – a specialized container engineered to optimize the
decomposition of organic materials into nutrient-rich compost.
8. 12V Power Supply – a device that converts mains AC (Alternating Current) electricity
from a wall outlet into a stable DC (Direct Current) voltage level.
9. NPK Testing Kit – is a soil testing tool measuring levels of Nitrogen (N), Phosphorus
(P), and Potassium (K). By analyzing these nutrient levels, the NPK Testing Kit helps
optimize fertilization strategies, a vital aspect of our bio-composter system aimed at
enhancing the efficiency and productivity of organic waste decomposition through
intelligent nutrient management.
10. Jupyter Notebook – is an essential open-source web application utilized within our
project for interactive data exploration, visualization, model development, and the
deployment of genetic algorithms. It serves as a critical tool for processing the data
1475

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

acquired by our bio-composter system, facilitating efficient model development and
analysis.
Design Framework of the Bio-Composter System
The researchers utilize a pre-assembled composter system as the foundational
framework for their experimentation. The pre-built composter system functions as the
foundation upon which further parts are easily added to improve efficiency and usefulness. The
addition of a stepper motor and the device driver is one such modification. The connection
between the stepper motor driver and the Arduino Mega board facilitates the comprehensive
control of the composter's rotational dynamics, including both its overall rotation and the
specific revolutions per minute (RPM) at which it operates.

Figure 1. Rotary Composter Bin Design

Figure 2. Overall view of Rotary Composter Bin
The system is integrated with sensors to monitor the temperature, humidity, Carbon
Monoxide content, Hydrogen content, and Ammonia Content. The sensors are then interfaced
with Arduino. The serial monitor of the Arduino is then interfaced with the Raspberry Pi to
display the values generated from the sensors.
The integration of a Hall effect sensor also adds another layer of precision and
convenience to the composting process. It makes the composter dock positioned for extraction
and insertion of the compost. That prepares the user to extract or insert compost material
efficiently.
1476

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Software and Hardware Interfacing of the Bio-Composter System

Figure 3. Bio-Composter System Flow Chart
The process initiates when the user selects an input through the GUI interface. If the
user selects an RPM, this command will be transmitted to the Arduino Mega which initiates
the motor. Upon activation, the sensor transmits the values through the Arduino Mega to the
GUI interface. Should the user opt to halt rotation, the sensor verifies the Hall Sensor and halts
the motor accordingly.
GUI Design and Remote Functionality of the Bio-Composter System
The GUI shows a selection of 3 RPM settings: 4, 6, and 8 RPM. Once the motor
activates, the timer will start as well. The timer will not rest until the stop motor command is
initiated. Thus, even if the user adjusts the RPM setting during the composting process, the
timer will not be reset. The sensor values for the Carbon Monoxide, Hydrogen, Ammonia,
Temperature, and Humidity are also displayed in the interface.
Additionally, to make full use of the Raspberry Pi 4, a remote function is added as a
feature. This will enable the user to access the GUI interface which makes the composter
control and monitoring system accessible from anywhere in the world if there is an internet
connection.

Data Gathering
Twenty-seven (27) samples of compost were produced which had varying properties
for operational parameters and materials. The parameters varied with the composter are the
RPM and the total runtime of the composter rotation. Altering the RPM refers to the speed at
which the compost materials are mixed and aerated, which is essential for the microbial activity
that breaks down organic matter. The variable RPM settings for the composter are 4, 6, and 8
RPM. The total runtime refers to how long the compost is subjected to the composter for
rotation. 1, 2, and 3 hours are the runtime variables used in this study.

1477

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

For the compost materials—soy pulp, vegetable peelings, and coffee grounds—the
composition ratio varies across different samples. For each sample, one compost material is
used with a different RPM and runtime variable. 27 combinations are made with each material,
RPM, and runtime which makes up the 27 samples for the dataset. After extraction, each
compost sample undergoes analysis using an NPK Testing Kit. The validity of this kit is
confirmed by comparing its results for two baseline samples with those obtained from
comprehensive laboratory NPK measurements. Furthermore, the researchers incorporated
assigned values for the Low, Medium, and High indicators of the chart; 3 for Low, 6 for
Medium, and 9 for High. Middle values between the indicators were also assigned for more
data diversity.
Development of the Genetic Algorithm Models
a. Data Source and Preprocessing
The study utilized a dataset compiled from various composting trials conducted by the
researchers. This dataset includes the operational parameters: Rotations Per Minute (RPM),
Total Runtime; and the types of materials used (Vegetable Peelings, Soy Pulp, Coffee
Grounds). The dataset also includes the resulting NPK (Nitrogen, Phosphorus, Potassium)
values for each trial.
b. Parameter Specification
The parameters and their respective ranges were specified as follows, allowing the
genetic algorithm to manipulate these within defined limits:
RPM: 4, 6, 8
Total Runtime: 1, 2, 3, 4, 5 hours

−
−

The target for optimization was set to achieve NPK values of 9, 9, 9.
c. Genetic Algorithm Configuration
Each material type will have a respective genetic algorithm model to determine the
best operational parameters for each material.
Each of the genetic algorithms is configured with the following components:
i.

Fitness Function:

The fitness is calculated based on the Euclidean distance from the target NPK
values, employing a squared error penalty to intensify the impact of larger
deviations:

1

1

,

ℎ

⬚

!,

"

# $%&# '

(

1478

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Figure 4. Fitness Function Computation
ii.

Population Initialization:

An initial population of 50 individuals was generated, with half derived from the
existing dataset of each respective compost material to ensure a realistic starting
point, and half randomly generated based on the defined parameter ranges to
introduce diversity. This approach is designed to enhance the algorithm’s
performance, compensating for the limitations of data acquired from testing kits with
restricted trial counts. By simulating potential genetic variations, the algorithm
expands the dataset, creating scenarios that mimic realistic genetic configurations
without the need for additional real-world testing.
iii.

Selection Mechanism:

Tournament selection was employed with a tournament size of 5, selecting the best
out of a randomly chosen subset of the population to proceed to reproduction.
iv.

Crossover and Mutation:

- Crossover: A two-point crossover method was used, where two random
crossover points are selected, and the segments between these points are swapped
between two parent individuals to create two offspring.
- Mutation: An adaptive mutation rate starting at 5% was applied, where
individual parameters of the offspring were randomly altered within their specific
ranges. The mutation rate was increased by 5% every 10 generations if no
improvement in fitness was observed.
d. Algorithm Execution
Each of the genetic algorithms ran for 100 generations, with each generation consisting
of the selection, crossover, and mutation processes to create a new population. The best
individual from each generation was tracked, and its fitness was compared to that of the best
individuals from previous generations.
e. Genetic Algorithms Output
Upon completion of the genetic algorithms, the optimized operational parameters of the
best individual for each compost material were output along with its respective fitness score.
These individuals represented the optimal set of composting parameters that were most likely
to achieve the target NPK values according to the model for each respective compost material.

1479

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Figure 5. Genetic Algorithms Flowchart

RESULTS AND DISCUSSION
Data Gathering Results
1. Coffee Grounds

Figure 6. Coffee Grounds Nitrogen Results
The x-axis represents the composter runtime values of 1-3 hours, and the y-axis shows
the resulting Nitrogen values of the compost upon being tested in the NPK testing Kit. Results
show that using coffee grounds as compost and running it at 6 Revolutions Per Minute for 3
hours results in high Nitrogen values which is the desirable amount for the target compost.

Figure 7. Coffee Grounds Phosphorus Results

1480

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Phosphorus values reach the desirable Phosphorus value of 8 when composting it at 6
rpm for three hours.

Figure 8. Coffee Grounds Potassium Results
The data shows that the target value of Potassium was reached by running the
composter for 2 hours on 8 rpm.
2. Soy Pulp

Figure 9. Soy Pulp Nitrogen Results
The x-axis represents the composter runtime values of 1-3 hours, and the y-axis shows
the resulting Nitrogen values of the compost upon being tested in the NPK testing Kit. Results
show that by using soy pulp, the highest value of nitrogen of 7 was achieved when the
composter was run for 3 hours at 4 rpm.

Figure 10. Soy Pulp Phosphorus Result
The experimentation shows that the optimal value of 8 was achieved by composting the
soy pulp in 8 RPM for 2 hours.
1481

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Figure 11. Soy Pulp Potassium Results
The graph shows that the target value of 9 was reached when the soy pulp is
composted with 6 rpm and an active run time of 3 hours.
3. Vegetable Peelings

Figure 12. Vegetable Peeling Nitrogen Results
Results show that a nitrogen value of 7 was reached given that the composter is running
at 6 Revolutions Per Minute for 3 hours.

Figure 13. Vegetable Peelings Phosphorus Results
The graphs show that a value of 9 for potassium is achieved given that the composter
is running at 6 rpm for 3 hours.

1482

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Figure 14. Vegetable Peelings Potassium Results
The graph shows that a value of 7 for potassium is achieved given that the composter
is running at 8 rpm for 2 hours.
b. Genetic Algorithm Models Fitness Testing
1. Coffee Grounds
Table 1
Genetic Algorithm Results of NPK using Coffee Grounds with Best Individual Results.
Target Values

Best Individual Parameters

Nitrogen: 9
RPM: 6
Potassium: 7
Total_Runtime: 3.0 hours
Phosphorus: 8
Coffee_Grounds: 1.0
Fitness: 0.1667

The table shows the best runtime and RPM results for the coffee ground material.
Nitrogen shows the highest value, reaching the desired target of 9, followed by Phosphorus
with a value of 8, and Potassium with 7. This result correlates with the dataset as coffee grounds
have high values for Nitrogen. However, the resulting compost parameters are already
indicated in the dataset. Additionally, the resulting NPK values from the dataset are not far off
from the NPK values of the genetic algorithm.
2. Soy Pulp
Table 2
Genetic Algorithm Results of NPK using Soy Pulp with Best Individual Results.
Target Values
Nitrogen: 8

Best Individual Parameters
RPM: 8.0

1483

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024
Potassium: 9

Total Runtime: 4.0 hours

Phosphorus: 8

Soy Pulp: 1.0

Fitness: 0.3333

Results for the soy pulp show that Potassium has the highest value, followed by
Nitrogen and Phosphorus having the same values. The best runtime and RPM for the composter
are 8 and 4 hours, respectively. The resulting runtime hours is 4 due to the adjustment of the
parameter ranges for the genetic algorithm, which includes 4 and 5 hours. Soy pulp has a higher
fitness score than coffee grounds.
3. Vegetable Peelings
Table 3
Genetic Algorithm Results of NPK using Vegetable Peelings with Best Individual Results.
Target Values

Best Individual Parameters

Nitrogen: 9

RPM: 6.0

Potassium: 9

Total_Runtime: 5.0 hours

Phosphorus: 8

Vegetable_Peelings: 1.0

Fitness: 0.5

NPK Results for the Vegetable peelings show the highest fitness score. 9, 9, and 8 for
Nitrogen, Potassium, and Phosphorus, respectively. Optimal compost parameters are 6 RPM
and 5 hours runtime. This indicates compared to the data gathered in the dataset the NPK values
should still rise with increased runtime hours.
c. Genetic Algorithm Fitness Analysis
All the genetic algorithms result in a rapid convergence in the fitness scores after a
certain generation. A range of factors are to be considered due to this. Firstly, the diversity of
the dataset, as there are only a limited number of samples collected for the data gathering due
to the test kit only having a limited number of trials. Additionally, the temporary constraints
of not being able to conduct runtime tests for data gathering reach up to a day or a week.
Thus, regular composting runtimes have not been fully emulated for the data gathering. The
researchers have already implemented diversity mechanisms by fine-tuning the function
parameters of the genetic algorithm and used artificially generated initial solutions to achieve
these results. Fitness functions have been modified as well, but results have shown similar or
worse outcomes. Finally, all these genetic algorithm models use a specific random seed for
1484

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

fair results. Thus, the results for each material also have varying NPK and compost parameter
values.

CONCLUSION AND RECOMMENDATIONS
Conclusion
In the design and optimization of a Bio-Composter System using Genetic Algorithm,
the researchers have the following conclusions:
−

−

−

The genetic algorithm achieves realistic results in determining the best composter
operational parameters for each compost material with the given target NPK
values. Rotten vegetables material has the highest fitness score, and coffee
grounds has the lowest. As a model suited for optimization problems, it uses
global and local search to find solutions for the best composting parameters that
are not present in the given dataset.
Composter system software provides ease of use through an interface for
composter control and monitoring systems. Remote capabilities also allow for
convenient handling and visual representation of data. Also, users can control the
system from anywhere virtually using the remote capabilities of the system.
The motor automation of the composter system ensures consistent rotation and
variable preset rotation speed.

Recommendations
1. Enhance Data Collection and Automation
-

Integrate advanced automation techniques, including automated extraction and
cleaning, to improve system efficiency, accuracy, and sanitation while minimizing
manual intervention.

2. Improve Hardware Performance
-

Extend the runtime durations of the Bio-Composter System to support continuous
operation.

-

Upgrade motor power and adjust gear ratios to handle varying loads effectively,
ensuring optimal torque and speed for enhanced composting performance.

3. Improve Genetic Algorithm Exploration
-

Implement and refine genetic algorithms for real-time monitoring and adjustment
of composting parameters.

-

Add more controllable parameters to improve the model results, leading to more
reliable and robust composting processes.

1485

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

4. Compost Material and Process Innovation
Investigate other compost types and production processes, including liquid compost
and preprocessing techniques.

1486

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

REFERENCES
(BSWM), B. o. (2021). Enhancing Compost Quality through Technology Integration. BSWM
Journal of Soil Management.

(IRRI), I. R. (2021). Innovative farming solutions to increase productivity in Philippine rice
farming. Retrieved from https://ati2.da.gov.ph/ati-car/content/article/bongbong-l-bulie/enhancing-productivity-through-innovative-rice-farming-techniques-aews

Agriculture, P. D. (2020). Philippine Agriculture in 2020: Status, Challenges, and Policies.
Retrieved from https://pidswebs.pids.gov.ph/CDN/PUBLICATIONS/pidspn2112.pdf

Authority, P. S. (2020). Fertilizer statistics in the Philippines, 2018. Philippine Statistical
Reports on Agriculture and Fisheries.

Chun-Fang Shi, H.-T. Y.-T.-P.-Y.-B.-G.-M. (2022, August). Optimization of Humic Acid
Content in GLR Composting. Bioresource Technology, 357, 127248.

Composting. (2014, February 3). Retrieved from Town of North Castle. NY:
https://www.northcastleny.com/sites/g/files/vyhlif3581/f/file/file/composting.pdf

Crisostomo, J. (2018). Economic impacts of high fertilizer prices in the Philippines. Journal
of Southeast Asian Economies, 35(3), 355-367.

David, C. C., & Inocencio, A. B. (2019). The state of Philippine agriculture: Challenges for
sustaining productivity growth. Philippine Review of Economics, 56, 95-122.

Grell, M., Barandun, G., Asfour, T., Kasimatis, M., Collins, A. P., Wang, J., & Güder, F.
(2021). Point-of-use sensors and machine learning enable low-cost determination of
soil nitrogen. Nat Food 2, 981-989.

1487

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Guochao Ding, X. S. (2023). The Influence of a CGA-BP Neural-Network-Based Aeration
Oxygen Supply Prediction Model on the Maturity of Aerobic Composting. Processes,
11(6), 1591.

Kim, M.-S., Kim, S.-C., Yun, S.-G., Park, S.-J., & Lee, C.-H. (2018). Quality Characteristics
of Commercial Organic Fertilizers Circulated. Journakl of the Korea Organic
Resources Recycling Association, 26(1), 21-28.

Kumar, S., Bharti, A., & Kumar, R. (2020). Composting parameters and compost quality: a
literature review. Environmental Technology & Innovation, 2020, Article 101026.
doi:10.1016/j.eti.2020.101026

Li, M., Li, S., Chen, S., Meng, Q., Wang, Y., Yang, W., . . . Guo, X. (2023). Measures for
Controlling Gaseous Emissions during Composting: A Review. International Journal
of Environmental Research and Public Health, 20(4), 35-87.

Malik, A. &. (2023). Advances and applications of genetic algorithms in machine learning
for materials science. Materials Today: Proceedings, 62, 6692-6698. .

Schmitt, J., Bönig, J., Borggräfe, T., Beitinger, G., & Deuse, J. (2020). Predictive modelbased quality inspection using Machine Learning and Edge Cloud Computing.
Advanced Engineering Informatics, 45.

Schorr, S., Möller, M., Heib, J., Fang, S., & Bähre, D. (2020). Quality Prediction of Reamed
Bores Based on Process Data and Machine Learning Algorithm: A Contribution to a
More Sustainable Manufacturing. In Procedia Manufacturing (Vol. 43, pp. 519-526).

1488