.jpg)
Phonlamai Photo / Shutterstock
In 2015, researchers at Carnegie Melon University and researchers at Intelligent Fiber Optic Systems Corp., along with support from NASA, developed a three-fingered robotic hand with fiber optic sensors embedded into the soft robotic skin. A total of 14 fibre optic sensors were incorporated into the robotic hand with the capability of detect force of contact of as little as less than a tenth of a newton. The sensors allow the hand to determine where its fingertips are and feeds back information on contact. The team also developed a new kind of stretchable optic sensor that would, in theory, be embedded into the robotic skin in order to provide an even greater level of feedback.
What’s important about this advancement in robotic hand technology is that it brings them one step closer to mimicking the qualities of human hands. There are thousands of tactile sensory units in just one human fingertip, but in contrast, NASA’s most state-of-the-art humanoid robot only has 42 sensors in total in its hand and wrist. The highly sensitive sensory system in the human hand allows it to behave in uniquely important ways. It allows us to react rapidly and safely to unexpected stimuli, while it allows us to be dextrous and accomplish tasks that require intricate, coordinated hand movements. In a very basic view, it also allows us to handle 3D objects appropriately, we can adjust our grip and grasp to execute precise manipulation.
While industrial robots require only a limited number of sensors to carry out their jobs, there is a need to develop a different breed of robots that can routinely and safely interact with humans. These robots will require increased tactile attention and sensitivity, meaning numerous sensors incorporated into soft robotic skin.
Fiber optics are the future for giving robotic hands increased sensitivity. Alternative methods of increasing robotic sensitivity have too many drawbacks to make them feasible. A greater number of conventional pressure or force sensors could be incorporated within a robotic hand, but this would lead to problems such as interference from the electromagnetic devices incorporated into the robot. In addition, they are also prone to breaking. Fiber optics overcome both these issues, as a single optical fiber can comfortably contain multiple sensors, is impervious to electromagnetic interference, and is durable and resistant to breakage.
The sensors used by the Carnegie team work by measuring the shift in the wavelength of light that is reflected by the optical fiber in order to measure strain. The new sensors that the team developed aimed to overcome the limitations of conventional optical sensors that are not very flexible. Even the most flexible of them only stretch 20-25%, which is a problem when applying them successfully to a robotic hand that requires a range of complex movements.
Working alongside researchers at the University of Texas, the Carnegie team was able to invent a flexible, highly stretchable sensor from silicon rubbers. The innovative sensors incorporate reflective gold into the soft waveguides, and when the silicon is stretched, cracks appear in this reflective layer and light can escape. This loss of light is then measured, and a calculation of strain or other deformations is possible.
Recently, a team at Cornell University made advancements in this direction. In 2016, following on from the developments at Carnegie Melon, the team created curvature, elongation and force sensors using stretchable optical waveguides in a soft robotic hand.
The path for the future of robotic hands is clear, developments in fiber optic sensors is certain to open the possibility to creating humanoid robots with the capabilities of fine motor skills and a high sensitivity to touch which will make them suitable for applications where they are interacting with humans.
References and Further Reading
https://www.theengineer.co.uk/
Disclaimer: The views expressed here are those of the author expressed in their private capacity and do not necessarily represent the views of AZoM.com Limited T/A AZoNetwork the owner and operator of this website. This disclaimer forms part of the Terms and conditions of use of this website.