SeGreait

A Science Research

(Innovation)

Rhuda Mikahaela Niña V. Balansag

Ella Philline G. Ingeniero

Jessie Januar A. Dalauta

Marc Constantine C. Dy

Sam Alexis Flores

RESEARCHERS

Rodman Rick P. Adobas

ADVISER

Mrs. Amor O. Abando

School Principal
 TABLE OF CONTENTS

I. Introduction

Features and Specifications

Market Trends and Opportunities

II. Materials and Methodology

III. Results and Discussion

IV. Conclusion

V. Acknowledgements

VI. References/Bibliography
 INTRODUCTION

Features and Specifications

SeGreait is an automated trash bin that utilizes artificial intelligence

in manufacturing automated trash bins by letting them to perform

complicated tasks and adapt to climate change, eventually improving

waste management processes and sustainability efforts.

First, infrared sensors will process if there's any presence of waste.

The NodeMCU will instruct the ESP-32 cam to capture the specified waste

unless it was properly detected by the infrared sensors then it will repeat

the same procedure if not properly sensed. If captured properly the

NodeMCU will send the captured image to the server then run the ResNet-

152 CNN Algorithm. Thus, proceeds to the waste classification. If it's

biodegradable the NodeMCU instructs the servo which is used to turn the

separator(plywood) to 250 degrees likewise for the non - biodegradable

servo however only turns to 20 degrees. Then the procedure will repeat if

waste is captured

Market Trends and Opportunities

Separated trash bins for non-biodegradable and biodegradable

waste has been used for a long time, and it has been effective in

segregating waste. However, the researchers have observed that there

are people who would still throw their waste in the wrong bin due to their

laziness or lack of discipline. Improper waste management can pose a

threat to our environment and to our health.

Nowadays, the advancement of technology rises because of our

growing desire to apply nеw technologies in making things easier and

convenient for our busy lifestyle. Accordingly, technology changed the

way people live, from cars that can drivе themselves, smart cameras, and

buses with automatic doors (Simplilеarn, 2023). As technology keeps

advancing, the researcher came up with an idea to design and develop

SeGreait. Unlike the ordinary trash bins that we are currently using, this

smart trash bin can automatically segregate non-biodegradable and

biodegradable waste that are commonly found in school such as paper,

plastic bottles, etc. with the help of Artificial Intelligence (AI) and sensors.

Through this device, it would not only enhance the way of segregating

trash and increase the user’s convenience, but it can also help in

mitigating environmental issues and ensure the proper waste

management in places such as classrooms or offices.

In a research study by Afaque Sayyed,et.al (2020) entitled

“Automatic Waste Sorting Machine” uses pipes to segregate metals ,

plastic, and glass while for our study, it aims to segregate non-

biodegradable and biodegradable waste through tilting the wooden

trapdoor.
 Another research study by Henita Rahmayanti et al.in 2019 entitled

“Prototype Smart Trash Can for Implementation Smart Environment in the

Smart City based on Arduino and Android” in which they used Arduino to

detect the type of garbage that enters the trash to control the servo

motor to open the door according to the type of garbage- plastic waste

and iron waste. Whereas for our study, we used NodeMCU ESP8266 chip-

based microcontroller development board, since its very cheap as

compared to Arduino UNO and it can also be easily inserted into a

breadboard and test various circuit designs. Moreover, our study focuses

only in segregating biodegradable and non-biodegradable wastes that are

common within the school campus.

 MATERIALS AND METHODS

Materials

 NodeMCU V3 ESP8266 ESP-  30 cm Breadboard

12E Wi-Fi development board connecting Jumper Wires

Female-Female

 Infrared IR Proximity Sensor  30 cm Breadboard

connecting Jumper Wires

Male-Male

 NodeMCU ESP8266  30 cm Breadboard

Expansion Base Board connecting Jumper Wires

Male-Female

 ESP32-CAM-CH340 USB  ¼ Plywood

Serial to Bluetooth and Wi-Fi  9V Battery

Camera Development Board

 Solder Less PCB Breadboard  ½ Plywood

 2 SG90 Micro 9g Servo Nylon

Carbon fiber gears

 Invention design

Control Box
31.5 in

NON-BIODEGRADABLE BIODEGRADABLE
Compartment

31.5 in

Trapdoor
13.8 in
7.87 in 7.7 in

11.81 in

As seen in the sketch of the front view, the designed device is 31.5 in tall,

with a width of 31.5 in. The wooden trapdoor with a width of 7.7 inches

and has a length of 13.8 inches. It, is placed inside the trash bin that will

tilt to the left or to the right. According to the sketch, the control box of
the device is also placed inside. The control box is 7.87 inches in width

and 11.81 inches length. The control box contains the NodeMCU V3

ESP8266 ESP-12E Wi-Fi development board, the NodeMCU ESP8266

Expansion Base Board, and the connected wires of the sensors, servos,

and camera that the device uses.

Program Flow
 Figure 1. Program Flow of the System
The figure above shows the procedural design of the system. The

system starts with the infrared sensor checking for the presence of waste,

after that, the sensors will detect. Then, the NodeMCU will instruct the

ESP-32 CAM to take a picture and send it to the server. Lastly, the servo
will run the ResNet- 152 CNN Algorithm to classify the waste. If its

biodegradable, the server will instruct the servo to tilt 250 degrees, and if

not, it will tilt to 20 degrees.

Information flow

IR PROXIMITY
SENSOR

NodeMCU V3
SG90 Micro ESP-12E WiFi
Servo development Server
board ResNet-152 CNN

ESP32-CAM-
CH340 USB

Figure 2. Information flow of the System

Illustrated in Figure 2, our automatic trash bin sorter uses sensors

like infrared IR proximity sensors to detect incoming trash. These sensors

send data to a NodeMCU V3 ESP8266 ESP-12E Wi-Fi development board,

which then transmits it to an AI module within an ESP32-CAM-CH340 USB

Serial to Wi-Fi Camera Development Board. The AI analyzes the trash's

biodegradability and instructs the SG90 Micro 9g Servo Nylon Carbon fiber

gears to tilt 250 degrees to the left/right. Based on the AI's decision, the
servo guides non-biodegradable waste in one direction and biodegradable

waste in another, enabling automatic sorting.

Architectural Design

Figure 3. Architectural Design of the System

Figure 3 presents the architectural design of the system. The

NodeMCU-V3 will serve as the main Microcontroller of the System. It is

responsible for data processing and operating the infrared sensor

checking for the presence of waste. Then it will instruct the ESP-32 CAM to

take a picture and send it to the server. The server will run the ResNet-

152 CNN Algorithm to classify the waste. If its biodegradable, the server
will instruct the SG90 Micro servo to tilt 250 degrees, and if not, it will tilt

to 20 degrees.

Methodology Flow Chart

 Finetune the AI using a dataset of images
Set up the SG90 Micro 9g Servo compiled by the research team if the
Nylon Carbon fiber Gears with AI Examine if the automated trash
accuracy and performance are not bin is successfully working.
sufficient

Performance of the developed device

The researchers will evaluate the performance of the

developed device by conducting experiments. The researchers will first

test whether the device would tilt to left if the thrown trash is non-

biodegradable and would tilt to the right if the thrown trash is

biodegradable. To do this, the device will be placed in a room within 5

days. After this, the researchers will record the occurrences where the

device successfully segregated according to the classification of the

waste. In addition, the researchers will get the percentage of each

occurrence where the device successfully segregated the non-

biodegradable/biodegradable waste set by the researchers.

User Instructions
 To use this invention, you must throw a single trash for the machine

to work. After dropping the trash, you must wait for the machine to

segregate your trash.

RESULTS AND DISCUSSIONS

 The automated trash bin will automatically tilt for 250 degrees to

the left if the waste is non-biodegradable and tilt for 250 degrees to the

right if the waste is biodegradable. The system will only process for one

trash at time.

WASTE TEST SEGREGATION TIME

Plastic Bottle Test 1 Unsuccessful 5 seconds

Test 2 Successful 11 seconds

Test 3 Successful 12 seconds

Test 4 Successful 11 seconds

Test 5 Successful 31 seconds

Plastic Wrapper Test 1 Unsuccessful 6 seconds

Test 2 Successful 11 seconds

Test 3 Successful 11 seconds

Test 4 Successful 11 seconds

Test 5 Successful 14 seconds

Table 1. Experimentation on Non-Biodegradable Wastes

Table 1 shows the data from the experimentation on non-

biodegradable wastes. Based on the given data, there are 80% of plastic

bottles that were properly segregated for 11–31 seconds, while 20% of
them were improperly segregated for 5 seconds. Plastic wrappers were

also 80% successful.

By conducting the experiments, the researchers found out that 80%

of the occurrences were successful, while 20% of the test was

unsuccessful because the system misclassified the leaf as non-

biodegradable.
 WASTE TEST SEGREGATION TIME

Leaf Test 1 Unsuccessful 11 seconds

Test 2 Unsuccessful 11 seconds

Paper Test 1 Successful 5 seconds

Test 2 Successful 13 seconds

Test 3 Unsuccessful 11 seconds

Test 4 Unsuccessful 4 seconds

Tissue Test 1 Unsuccessful 11 seconds

Test 2 Successful 10 seconds

Test 3 Unsuccessful 11 seconds

Test 4 Unsuccessful 12 seconds

Carton Box Test 1 Successful 4 seconds

Test 2 Successful 11 seconds

Test 3 Successful 11 seconds

Table 2.Experimentation on Biodegradable wastes

Table 2 shows data from the experimentation on Biodegradable

wastes. Based on the given data, there are 0% success rate of leaf that
were properly segregated for 11 seconds. Papers were 50% successfully

segregated for 5-13 seconds and 50% unsuccessful. Tissue had a 25%

success rate for 10 seconds, while 75% rate of it were improperly

segregated for 5-11 seconds. The carton box were 100% success rate that

were properly segregated for 4-11 seconds.

Through conducting the experiments, the researchers found out

that 46.15% of the occurrences were successful while 53.85% were

unsuccessful because the AI misclassified that the biodegradable wastes

as non-biodegradable.
 Functionality

The researchers monitored the mechanism of the automated trash

bin to rate its functionality. With the experiments conducted, the

researchers concluded that the mechanism’s functionality is 100%. This is

if the error in the code is excluded.

Another way for the researchers to rate the device’s functionality is

to monitor whether the sensors work properly. With the experiments

conducted, the researchers observed that the prototype can properly

segregate the non-biodegradable and biodegradable waste.

 Recommendation

This research focused on improving automated trash bin to help the

government and other private sectors that are concern on environmental

protections through proper waste management.

The researchers reviewed the impact and results of said study to

come up with these possible recommendations.

1. By developing educational materials as campaign materials to raise

awareness about the importance of waste segregation and to

convince the community in using the automated trash bin.

2. Formulate a maintenance plan to ensure the automated trash bin in

good condition, including regular inspection and cleaning of the

trash bin.

3. By enhancing the infrared sensor for better range, investing to a

stronger server and internet connection to improve the AI’s

capability to segregate trash.

 CONCLUSION

Based on the experiments conducted, the researchers successfully

designed and developed automatic trash bin with a good-performing

functionality package with eco-friendly materials . The results have shown

that the developed device has 99% efficiency, and the performance of the

developed device is 80%.

Researchers applied advanced technologies such as sensors and AI

to enhance the efficiency and accuracy of waste separation. SeGreait

Project designed the automated trash bin to be more natural and easier to

use for all users, including those with disabilities or limited mobility.
 ACKNOWLEDGEMENT

First and foremost, we would like to express our sincerest gratitude

to the Almighty Father for giving us the strength and guidance to continue

our study.

In addition, we would also like to acknowledge the contributions of

Mr. Nick Anthony Miras and Sir Rodman Rick Adobas for imparting their

knowledge, time, and efforts. To Mr.Recson Magallano Rebaldo for

making our ideal prototype into reality. Lastly, we would like to thank our

family and friends for their never-ending support and encouragement.

 BIBLIOGRAPHY

Simplilearn, Simplilearn. Has Technology Improved Our Lives? | Simplilearn. Retrieved

from Simplilearn, 2023. https://www.simplilearn.com/how-has-technology-improved-
our-lives-article#:~:text=From%20communnstant%20messaging%20and
%20facetime.jication%20and%20transport%20to,to%20i.

Myles Joshua Tan. “(PDF) Automated Waste Segregation System Using

Trained Optical ...” Automated Waste Segregation System using
Trained Optical and Material Sensors with User Communication
Capabilities, 2020.
https://www.researchgate.net/publication/342747452_Automated_Wa
ste_Segregation_System_using_Trained_Optical_and_Material_Sensor
s_with_User_Communication_Capabilities.

Sayyed, Afaque. Automatic Waste Sorting Machine , 2020.

https://www.researchgate.net/publication/332056188_Prototype_Sma
rt_Trash_Can_for_Implementation_Smart_Environment_in_the_Smart_
City_based_on_Arduino_and_Android.

Rahmayanti , Henita. Prototype Smart Trash Can for Implementation

Smart Environment in the Smart City based on Arduino and Android,
2019.
https://www.researchgate.net/publication/332056188_Prototype_Sma
rt_Trash_Can_for_Implementation_Smart_Environment_in_the_Smart_
City_based_on_Arduino_and_Android.

Sarnaik, Apoorva Sarnaik. “WASTE CLASSIFICATION USING RESNET-152.”

WASTE CLASSIFICATION USING RESNET-152:, 2023.
https://www.researchgate.net/publication/366932747_WASTE_CLASSI
FICATION_USING_RESNET-15.

J. Bobulski, J. Piatkowski, PET waste classification method and Plastic

Waste DataBase WaDaBa, Advances in Intelligent Systems and
Computing, vol. 681, Springer Verlag, 2018, pp. 57-64