Smart Robotic Arm In Medicine Allocation Application

Ong, Toby Yi Kang (2019) Smart Robotic Arm In Medicine Allocation Application. Project Report. Universiti Teknikal Malaysia Melaka, Melaka, Malaysia. (Submitted)

Text (24 Pages) Smart Robotic Arm In Medicine Allocation Application.pdf - Submitted Version Download (1MB)

Abstract

Medications are the most typical treatment intervention that used in healthcare to improve the health and well-being of patients. However, medication errors contain potential risk in affecting patient safety as well as treatment costs which then cause hazards for patients and their family. Hence, this project has proposed a Smart Robotic Arm in Medication Allocation Application in that to assits healthcare profession by reducing their tasks on medication dispensing. This project is developed by using TeraSoft 6 Degree of Freedom (DOF) Intelligent Technology Robot Arm System with code label based object recognition model. The inverse kinematic model is designed in Simulink Environment to control the position of Smart Robotic Arm for medication allocation. While, the code label based object recognition model is designed with Raspberry Pi Camera Module V2 and OpenCV ran in Raspberry Pi Model 2B for reading information of desired medication. The performance of the Smart Robotic Arm system is tested. The results have showed that the designed inverse kinematic model is capable in providing accurate solution after comparing with kinematic model that built in Peter Corke Robotics Toolbox. However, the designed robot controller has showed the limitation in optimizing the performance since the Smart Robotic Arm have experienced the steady-state error as well as vibration issue during positioning. Hence, the poor performance of robot controller has lead to the errorneous of Smart Robotic Arm in reaching the desired position in actual environment. Besides, the physical defecting in Smart Robotic Arm has also directed to the errorneous in reaching the desired position as well. Other than that, the code label based object recognition is built successfully and it has reached 100% accurate in detection and recognition of clear and blurred Quick Response (QR) code in 5cm between camera sensor and QR code image.

Actions (login required)

View Item