Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM62833
Volume:
10 |
5 |
May 2026
Pages:
1-3
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
MECHATRONICS
SANJIVANI COLLEGE OF ENGINEERING
MECHATRONICS
SANJIVANI COLLEGE OF ENGINEERING
MECHATRONICS
SANJIVANI COLLEGE OF ENGINEERING
MECHATRONICS
SANJIVANI COLLEGE OF ENGINEERING
MECHATRONICS
SANJIVANI COLLEGE OF ENGINEERING
In modern manufacturing, ensuring product quality is essential for operational efficiency and customer satisfaction. This research presents an integrated bottle shape defect detection system in a conveyor line using machine learning and transfer learning (MobileNetV2). A bottle serves as a prototype model, demonstrating how the same setup can be adapted to detect defects in a variety of industrial components such as containers, mechanical parts, and packaging materials. The proposed model captures real-time video frames using a Raspberry Pi 4-based vision system and classifies the bottles as defective or fair with a trained deep learning model. A servo-activated rejection mechanism, synchronized with the conveyor, removes defective bottles based on the classification results. This system reduces manual inspection errors, improves accuracy and increases production throughput. Experimental results confirm the effectiveness of the model, achieving high accuracy in real-time scenarios. The developed framework is scalable, cost effective and can be adapted to other manufacturing lines. This research advances intelligent visual inspection by merging conveyor-based automation and machine learning for real-time defect detection in industrial applications.
KEYWORDS
This paper outlined the design and implementation of a machine learning-based bottle shape defect detection system integrated with a conveyor automation setup. The proposed framework successfully showed how transfer learning using MobileNetV2 can achieve high accuracy in real-time classification of defective and proper bottles. By combining computer vision, Raspberry Pi-based embedded computing, and servo-controlled actuation, the system automates quality inspection without manual intervention. Testing confirmed that the model performs efficiently, achieving over 95% accuracy during real-time classification while maintaining a quick response time suitable for continuous conveyor operation. The major advantage of this system is its scalability and adaptability. Although the project used a bottle as a prototype, the methodology can be applied to detect defects in any industrial product, such as packaging materials, plastic components, or metal parts, by retraining the model on specific datasets. This makes it an ideal solution for smart manufacturing, directly supporting the objectives of Industry 4.0. Future work will focus on enhancing the system’s robustness by incorporating multi-class defect categorization, edge computing for faster processing, and IoT connectivity for remote data logging and predictive analytics. Further, implementing robotic sorting arms or collaborative robots (cobots) could enhance automation efficiency in high-speed assembly lines. In conclusion, the developed system offers an intelligent, low-cost, and real-time defect detection solution that can significantly enhance industrial product quality, reduce human error, and optimize manufacturing throughput. This research lays the groundwork for scalable
[1] S. Patel, R. Sharma, and P. Singh, “Automated Visual Inspection System for Bottle Defect Detection Using Image Processing,” Int. J. Adv. Res. Electr. Electron. Instrum. Eng., vol. 9, no. 5, pp. 1234–1240, 2023. [2] A. Kumar and N. Gupta, “Real-Time Quality Inspection on Conveyor Using Deep Learning,” IEEE Access, vol. 11, pp. 45120–45128, 2023. [3] R. S. Dubey and M. Jain, “Machine Vision-Based Defect Detection for Industrial Applications,” Procedia Computer Science, vol. 192, pp. 3100–3108, 2021. [4] T. Brown and K. Patel, “Implementation of Raspberry Pi for Industrial Automation,” Int. J. Eng. Res. Technol. (IJERT), vol. 11, no. 2, pp. 214–219, 2022. [5] P. K. Singh and R. Mehta, “Design and Control of Conveyor Belt System Using Servo Motor and Microcontroller,” IEEE Trans. Ind. Electron., vol. 70, no. 8, pp. 7655–7664, 2023. [6] TensorFlow, “TensorFlow: An Open Source Machine Learning Framework for Everyone,” [Online]. Available: https://www.tensorflow.org/. [Accessed: Nov. 11, 2025]. [7] OpenCV, “Open Source Computer Vision Library,” [Online]. Available: https://opencv.org/. [Accessed: Nov. 11, 2025].
Publication Date
5/8/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
6
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...