Swiss Researchers Develop Walking Robotic Hand Inspired by 'Thing'
Researchers at ETH Zurich have unveiled a robotic hand capable of walking on its fingers, utilizing machine learning to navigate and perform tasks.
A team from the Soft Robotics Lab at ETH Zurich has introduced a novel robotic hand that can move independently by using its fingers as legs. The innovation, which has drawn comparisons to the character 'Thing' from the popular series 'Wednesday', is capable of walking, recovering from falls, and even typing on a keyboard. According to doctoral student Amirhossein Kazemipour, the primary engineering challenge was coordinating the distinct fingers to maintain balance and movement without a central body. Instead of traditional manual programming, the researchers utilized machine learning to train the device to synchronize its movements. While this development is currently a research prototype, it represents a significant advancement in soft robotics and autonomous movement. For residents and tech enthusiasts in Thailand, this highlights the rapid evolution of robotics that may eventually influence assistive technologies or automated services. It remains to be confirmed how this technology will be scaled for practical, real-world applications beyond the laboratory setting. The research team has shared their findings via a pre-print paper and demonstration videos, showcasing the hand's ability to navigate surfaces autonomously.
