Recently, a study from the Swiss Federal Institute of Technology in Lausanne (EPFL) made its debut at the IEEE International Conference on Robotics and Automation in Rotterdam. The research aims to explore how robotic hands can break through current limitations and grasp a wider variety of objects.
The research team noted that deep learning models have significantly enhanced the dexterity of multi-fingered hand manipulation. However, in cluttered environments, grasp guidance based on contact information remains under-explored. To address this, the researchers designed a hand inspired by humans, capable of bending backward to pick up various objects and even climbing onto another object.
The results of this study might be somewhat unsettling, as the appearance and behavior of the robotic hand differ significantly from traditional bionic robotic hands. Nonetheless, the technology has sparked broad interest in the field, as it could represent a new direction for robotic grasping technology.
Researcher Xiao Gao stated, "Although you might see it in horror movies, I believe we are the first to introduce this concept into robotics." The success of this research could witness the replacement of bionic robots by terrifying robots.