Seminars and Events at automatic control
All seminars are held at the Department of Automatic Control, in the seminar room M 3170-73 on the third floor in the M-building, unless stated otherwise.
Master Thesis Presentation by Timothy Stjernfeldt and Clara Ulmestrand: Vision-Based Feedback Control for High-Precision Beam Alignment using a Robotic Arm at ESS
Disputation
From:
2025-08-18 14:00
to
15:00
Place: Room M:J, M-building, LTH
Contact: bjorn [dot] olofsson [at] control [dot] lth [dot] se
Date & Time: August 18th, 14:00-15:00
Location: Room M:J, M-building, LTH
Author: Timothy Stjernfeldt and Clara Ulmestrand
Title: Vision-Based Feedback Control for High-Precision Beam Alignment using a Robotic Arm at ESS
Supervisor: Björn Olofsson (LTH), Karin Rathsman (ESS), and Tomasz Brys (ESS)
Examiner: Yiannis Karayiannidis (LTH)
Abstract: This master thesis presents the development of a vision-based feedback control system for automating sample placement in a simulated neutron beam using a laser. The project was carried out on behalf of the European Spallation Source (ESS) to investigate the feasibility, possibilities and limitations of using a robotic automation system for sample alignment. Due to the radioactive research environment at ESS, there is need for robotic sample handling to minimize risk for personnel as well as increase overall productivity.
The system consists of a Stäubli TX60 robotic arm, two Basler CMOS cameras, a laser source simulating a neutron beam, and a sample prototype. An object detection model based on YOLO was successfully trained to track the robot gripper, and an operating interface (OPI) was developed to monitor the system and integrate it with the control system used at ESS.
The completed system is capable of automatically positioning the sample in the simulated beam and adjusting its orientation with the use of computer vision and algorithms that convert diffraction and backscatter patterns into movement commands for the robot. It reaches sub-millimeter positional repeatability and an orientational accuracy of approximately 0.1 degrees.
These results show that a vision based robotic feedback system is a feasible and effective approach for automated sample alignment in research environments that demand high precision. The project has contributed to valuable insights into both the capabilities and limitations of such a system, laying a foundation for future development and integration into real ESS operations. To make this system usable in real experiments at ESS, future work will be needed to adapt the system to analyze neutron scattering patterns rather than the simplified diffraction and backscatter patterns used in this project.