Improvement of SCARA robot application in automated blood testing using centrifugation and image processing techniques

Abstract

This study proposes an automated system for the blood testing process, specifically the blood centrifugation step, applicable in biomedical engineering. The system reuses the mechanical structure of an existing 4-degree-of-freedom industrial SCARA robot arm while developing additional components, including a vacuum-suction gripper for test tubes, driver circuits for the robotic motors, and dedicated robot control software. Subsequently, the robotic system is integrated with an automated centrifuge and a multi-camera image processing subsystem to create a fully automated blood centrifugation analysis and sample classification platform. Image processing is performed using two camera systems in this study: the first camera system is employed to pick and place test tubes from a specialized tray into the centrifuge machine to separate blood into serum and hematocrit in preparation for the subsequent ‘blood test’ phase, and the second camera system is used to detect whether the post-centrifugation test tube sample meets the requirements and to associate the data with a patient information label via a barcode on the test tube. The research successfully developed an experimental system, achieving a blood centrifugation cycle time of under 10 minutes per batch and an overall system reliability of 95% during operation.