1.
Theme
Information and Communication Technology
2. Problem Statement
How might we develop a solution for connecting farmers with local markets, enabling
real-time price discovery, efficient supply chain management, and fair pricing for agricultural
produce.
3. College Code and College Name
College Code: 2116
College Name: Rajalakshmi Engineering College, Chennai
4. Guide Name, Designation, Mobile No. & Email ID
Guide Name: Dr.V.S Selvakumar
Designation: Professor/ Department ECE
Mobile No.: 8610340723
Email ID: selvakumar.vs@rajalakshmi.edu.in
5. Student Team Details
S.No. Student Reg. Name of the Branch Mobile No. Email id
No. Student
1 210801059 Harini N ECE 9003331144 210801059@rajalakshmi.edu.in
2 210801019 Archana D ECE 8122327925 210801019@rajalakshmi.edu.in
3 210801046 Dinesh K ECE 9176213038 210801046@rajalakshmi.edu.in
4 210801051 Gayathri S R ECE 8838753955 210801051@rajalakshmi.edu.in
6. Project Summary
This project introduces an automated sorting solution using a conveyor belt system integrated with IoT
and machine learning technologies. The system ensures efficient sorting and quality assessment of fruits
and vegetables based on freshness, size, and other quality parameters. Key components include a
Raspberry Pi for processing, a gas sensor for freshness detection, and a camera module for image
classification. The system further enhances traceability by generating QR codes for each categorized
batch, enabling end-to-end transparency and trust in the supply chain.
7. Proposed Solution with Methodology
Proposed Solution:
The solution integrates hardware and software to provide a scalable, automated sorting system:
1. Freshness Detection:
○ A gas sensor positioned at the entry point of the conveyor belt evaluates gas emissions,
identifying spoilage indicators like ethylene levels.
2. Image Classification:
○ A camera captures high-resolution images of produce for analysis by a TensorFlow Lite
machine learning model.
3. Sorting Mechanism:
○ Servo motors and relays automatically sort items into bins based on quality
classifications.
4. QR Code Generation:
○ A Python script generates QR codes containing detailed information such as freshness
score, harvest date, and size.
Methodology:
1. Hardware Setup:
○ Assemble a motorized conveyor belt with gas sensors, cameras, and servo mechanisms.
2. Model Training:
○ Train a MobileNetV2-based machine learning model to classify freshness and size.
3. Integration:
○ Deploy the trained model on a Raspberry Pi and integrate hardware components.
4. Testing and Optimization:
○ Conduct iterative testing to refine classification accuracy and sorting reliability.
5. Deployment:
○ Demonstrate a fully functional prototype with real-time sorting capability.
8. Work Plan / Time Schedule Indicating the Project Milestones
Phase Activities Timeline
Phase 1 Finalize requirements and system design Week 1-2
Phase 2 Assemble hardware and integrate sensors Week 3-4
Phase 3 Train machine learning model and deploy on Raspberry Pi Week 5-6
Phase 4 Develop and integrate software for QR code generation Week 7
Phase 5 Test the system with sample data Week 8
Phase 6 Document and present the prototype Week 9
9. Plan of Action for Implementation
1. Assemble the conveyor belt system with sensors and actuators.
2. Train and deploy the machine learning model for image classification.
3. Test and refine the sorting mechanism using real-world data.
4. Integrate QR code functionality for traceability.
5. Conduct validation tests to ensure scalability and reliability.
10. List of Facilities Available in the College to Develop the Prototype of the Project
● Embedded Systems Lab: Raspberry Pi, gas sensors, servo motors, and relays.
● AI/ML Lab: Workstations for training machine learning models.
● Prototyping Lab: Tools for hardware assembly and fabrication.
● Faculty Expertise: Guidance in IoT, machine learning, and embedded systems.
11. Nature of Industry Support for the Project
● Collaborate with local farms or agricultural supply chains for real-world testing.
● Seek partnerships with IoT and ML solution providers for technical support.
● Approach hardware manufacturers for component sponsorship.
12. Total Cost
Item Cost (INR)
Raspberry Pi 5 with accessories 5,000
Gas Sensor Strip 1,000
Servo motors and load cell 1,500
Miscellaneous (camera, conveyor belt) 2,500
Total 10,000
13. Details of Financial Assistance Required
A financial grant of Rs. 10,000 is requested to cover the costs of sensors, Raspberry Pi, servo motors, and
prototyping expenses.
14. Expected Outcomes / Results
● Efficient sorting of produce with 95% classification accuracy.
● Enhanced traceability through QR code generation.
● Scalable, cost-effective solution for large-scale agricultural applications.
UNDERTAKING
1. ALL the students are studying in final year engineering. All the students are registered
only once for this scheme.
2. The college will provide the basic infrastructure and other required facilities to the
students for timely completion of their projects.
3. The college assumes to undertake the financial and other management responsibilities
of the project. We are aware that the amount is to be utilized only for the purpose
sanctioned i.e. to meet the expenses for developing the prototype and not for purchase
of computer consumables, stationeries, honorarium, overhead etc. Unutilised balance
amount will be returned back to the University after the time of completion of the project.
Name and Sign of Name and Sign of Name and Sign of Name and Sign of
Student 1 Student 2 Student 3 Student 4
Signature of the Mentor Signature and seal of the principal
