A control technology has been developed that can mitigate severe vibrations and unstable movements that occur when a robot arm in an industrial setting suddenly lifts a heavy object or receives an external impact.
A research team led by Kang Sang-hoon, a professor in the Department of Mechanical Engineering at Ulsan National Institute of Science and Technology, said on the 26th that it developed an "adaptive PID (proportional-integral-derivative) control algorithm" that helps a robot arm stably follow a target trajectory even amid rapid load changes and external contact.
A PID controller is a core device that calculates in real time the force a motor needs so a robot arm can execute planned movements precisely. Because it has a simple structure and is easy to apply, most industrial robot arms use this control method.
However, conventional PID control operates based on initially set control values, so when conditions change—such as when the weight of a lifted object changes or when there is contact with an external object—errors can grow, leading to vibrations or malfunctions.
To address this weakness, the team proposed an algorithm designed to let the robot adjust control values on its own by using error information that arises during operation. When environmental changes are detected, the robot updates control parameters in real time to reduce oscillation and maintain the desired trajectory.
In particular, a strength of this technology is that it can be applied by updating only the software without changing the robot's hardware. The team said that if a robot already has a PID controller, it can be adopted without additional equipment.
The team verified performance by applying the algorithm to a robot arm with two joints. In experiments, the robot arm stably tracked the target trajectory by adjusting control values on its own even when the work environment changed, and it maintained motion without vibration.
Kang Sang-hoon said, "It has strong potential for use not only in Smart Factory environments where working conditions change frequently, but also in various fields such as rehabilitation robots and Humanoid Robot that must respond to subtle changes in human force."
The findings were published on the 13th (local time) in IEEE/ASME Transactions on Mechatronics, an international journal in mechanical and robotics engineering.
References
IEEE/ASME Transactions on Mechatronics(2026), DOI: https://doi.org/10.1109/TMECH.2025.3634620