Automation of a PCB teststation with UR3 - An Applied Study in Collaborative Robotics and Test Process Engineering

This thesis investigates how a UR3 robot can automate PCB testing in a lab environment to improve efficiency and reduce manual workload. The results show that while a built-in control system offers high repeatability, a more advanced setup using external programming enables greater flexibility, highlighting both the potential and challenges of robotic automation in lab-scale testing.

Read the full thesis

Authors:  Alexander Forslind & Alexander Smith

Examiner: Björn Johansson

Supervisor: Hao Wang

Co-supervisor: Kristoffer Lagerström

Year: 2025

Manual testing of printed circuit boards (PCBs) is a time consuming and repetitive process that can lead to ergonomic strain and inefficient productivity. This thesis explores how collaborative robots can be used to automate the PCB testing process in a lab scale environment. The objective was to design, build and evaluate a prototype test station using the UR3 collaborative robot to perform automated pick and place, handling, test initiation and sorting of PCBs.

The system was developed in two parallel configurations. The first used the UR3s built in interface called Polycope and interacted directly with the other components through the robots I/Os. The second setup combined a Raspberry Pi with Python programming and Real Time Data Exchange (RTDE) to enable external control, vision based positioning and more dynamic system behavior. A fixed camera mounted above the conveyor was used to implement edge detection for identifying the PCB's orientation while capacitive sensors and PWM-regulated motors supported component detection and transport.

To support the physical integration of components, a number of custom designed mechanical parts were prototyped using 3D printing. These included gripper mounts, sensor housings, camera holders, claws and sorting trays, all tailored to match the layout and functional needs of the station.

Testing showed that the UR3 robot could reliably automate the core steps of the test process. The PolyScope-based system provided high repeatability under fixed conditions, while the Python-based setup allowed for greater flexibility and modularity. Some challenges remained in the implementation of vision-based positioning, particularly in terms of distinguishing the PCB from the conveyor background. However the overall system fulfilled the defined functional goals and demonstrated the possibility of using a collaborative robot to improve efficiency, reduce physical workload and increase consistency in test workflows

Read the full thesis