In context
By early 2024, human-machine interfaces for robotic arms were still limited by sparse command sets and cumbersome operation, which kept them from practical multi-dimensional control. Assistive and rehabilitation robotics needed more intuitive, reliable ways for users—especially those with upper-limb impairments—to direct a robot through continuous motion.
What was reported
Researchers at South China University of Technology and Pazhou Lab proposed a wearable hybrid interface that fuses electrooculography (EOG), head posture and speech into control commands. The system supports continuous 2D and 3D motion of a robotic arm at arbitrary angles, addressing the limited-command problem of earlier single-modal interfaces.
Ten subjects performed command output, 2D target tracking, letter writing and 3D object grasping tests. Blink-generated commands achieved an average accuracy of 96.67%, a response time of 1.51 s, an information transfer rate of 142.53 bit/min and a false positive rate of 0.05 events/min. In 2D tracking, the normalized root mean square deviation along two routes was 0.12 and 0.14; in 3D grasping, the average trajectory efficiency reached 92.65%.
The authors reported that control performance was comparable to manual operation, and statistical correlation tests supported the interface's reliability for continuous arm guidance.
Why it mattered
By combining multiple natural signals in a wearable package, the work pointed toward more practical assistive interfaces that could give users with limited mobility finer, more continuous control over robot arms—an important step for rehabilitation and home-use automation.
“The control performance of the system is comparable to manual control.”
Source: 《机器人》期刊 (robot.sia.cn) · Published 2024-02-04 · “基于可穿戴式多模态人机接口的机械臂运动控制方法”