This research aims to develop an automatic gemstone color sorting machine to overcome the limitations of manual color sorting, which can be restricted by speed and accuracy. This study applies deep learning technology to analyze and classify gemstone colors precisely, developing an algorithm capable of accurately detecting and categorizing color shades. An automated conveyor system was also designed to efficiently transport gemstones through the color sorting process, allowing for continuous operation. The sorting machine works by capturing high-resolution images of the gemstones, processing them with software to classify color shades, and directing each gemstone to its designated position on the automated conveyor. Experimental results demonstrate that the automated color sorting machine, integrated with the conveyor system, achieves high speed and accuracy, significantly reducing labor costs and enhancing the efficiency of gemstone color sorting.
ปัจจุบันการตรวจสอบเฉดสีของพลอยถือเป็นขั้นตอนสำคัญในการประเมินคุณภาพ ทั้งในกระบวนการผลิตและการค้าขาย อย่างไรก็ตามการตรวจสอบเฉดสีที่ดำเนินการโดยมนุษย์นั้นมีข้อจำกัดหลายประการ ซึ่งหนึ่งในนั้นเป็นเรื่องความสามารถในการแยกเฉดสีที่ซับซ้อน ซึ่งอาจทำให้การตรวจสอบใช้เวลานาน และมีความแม่นยำต่ำ จากปัญหาดังกล่าว จึงได้พัฒนาแนวคิดในการสร้างเครื่องคัดแยกเฉดสีพลอยอัตโนมัติ โดยใช้ระบบ Computer Vision ร่วมกับระบบอัตโนมัติในการวิเคราะห์เฉดสีของพลอย เพื่อลดข้อจำกัดของการตรวจสอบด้วยมนุษย์ เพิ่มความแม่นยำและประสิทธิภาพในการทำงาน รวมถึงทำให้กระบวนการตรวจสอบเป็นไปอย่างรวดเร็วและมีมาตรฐานของเฉดสี GIA (Gemological Institute of America)
คณะวิศวกรรมศาสตร์
This research project focuses on the design and development of a Manual Control Robot using Load Cell technology to enhance precision and reduce the time required for robot control. The use of automation robots in industries still presents challenges due to the complexity of programming and control. Therefore, developing a manual control system that responds to force input in all directions can significantly improve the efficiency of robots, making them more suitable for tasks requiring precise and intricate control. The study integrates Load Cell sensors, an HX711 amplifier circuit, and an Arduino UNO R3 to develop a control module that translates user-applied forces into commands for an RV-7FRL-D industrial robotic arm. Additionally, MATLAB is utilized for processing Load Cell data to analyze and optimize the robot’s movement accuracy. The results demonstrate that the developed system effectively reduces robot setup time while simplifying and improving control flexibility. This project represents a crucial step in enhancing the capabilities of industrial robots, allowing for seamless human-robot interaction through a manual control system that directly responds to user-applied forces.
คณะวิศวกรรมศาสตร์
An automated hydroponic system for household use has been developed to cater to individuals with limited space who wish to conveniently and easily grow their own salad greens at home. This system is designed to automatically control nutrient delivery by setting appropriate electrical conductivity (EC) and pH levels tailored to the specific salad greens being grown. It includes artificial lighting to enable cultivation in confined spaces with insufficient sunlight and is more cost-effective than similar systems available on the market. System testing revealed that the automated control of EC and pH values performed effectively, achieving the preset levels within 30 minutes and maintaining them consistently throughout operation. In an experiment growing green oak lettuce using a simulated balcony setup, the plants demonstrated a higher growth rate compared to conventional methods, particularly when artificial lighting was used.
คณะเทคโนโลยีการเกษตร
Expanding from a public park design project to a campus design on an area of over 50 rai in Ang Sila Subdistrict, Mueang District, Chonburi Province, to serve as both an educational institution and a place for relaxation and learning for the surrounding people.