Improving Human-Robot Interactions with ProTac: A Soft Robotic Link with Tactile and Proximity Sensing Capabilities
In recent years, robots have become highly advanced machines capable of assisting humans in various tasks. They are no longer confined to working behind security barriers, as they can now operate alongside humans in close contact. However, it is crucial to ensure that these robots are safe and pleasant for humans to interact with. This is where the development of ProTac, an innovative soft robotic link with tactile and proximity sensing capabilities, plays a significant role.
The Importance of Sensing in Human-Robot Interactions
Sensing the environment is crucial for robots to interact effectively with humans. While robots can use touch, sound, and sight to perceive their surroundings, tactile sensation is particularly important when coming into physical contact with humans. Unfortunately, the development of large-scale tactile sensors has faced numerous challenges, and most research has focused on responding to physical touch rather than touchless stimuli. To address these issues, a research team from Japan Advanced Institute of Science and Technology (JAIST) has developed ProTac, which combines tactile and proximity sensing capabilities.
The Innovative Design of ProTac
ProTac is a soft, cylindrical segment designed to be incorporated into a robotic arm. It features an outer “soft magic skin” that can withstand deformation without damage. Inside the skin, arrays of reflective markers are patterned, and fisheye cameras are installed at both ends of the robotic link. When the skin is deformed upon physical contact, changes in the markers’ positions are captured by the cameras. These changes are then processed to calculate the precise location and intensity of the contact. Additionally, the outer skin is made of a functional polymer that can be made transparent, allowing the cameras to capture footage for proximity calculations.
The Simulation and Learning Framework for Training ProTac
To train ProTac in making proximity and tactile measurements, the research team developed SimTacLS, an open-source simulation and learning framework. Built on the SOFA and Gazebo physics engines, this machine learning framework considers the physics of soft contact and the realistic rendering of sensor images. Using simulated and experimental data, the framework enables effective implementation of tactile perception in robotic links without the need for costly and complex experimental setups. It also allows users to validate sensor designs and learning-based sensing performance before fabrication and implementation.
In conclusion, the development of ProTac and the SimTacLS framework opens up new possibilities for improved human-robot interactions. By incorporating tactile and proximity sensing capabilities, robots can operate safely and predictably when in close contact with humans. This advancement paves the way for a future where humans and robots can work harmoniously together. The techniques developed can be extended to other types of robotic systems, such as mobile and flying robots, enabling robotic manipulation in cluttered environments or when operating near humans.