In a ground-breaking development, a team of researchers at the University of Cambridge has designed a soft, 3D-printed robotic hand capable of performing complex movements. The unique feature of this robotic hand is that it does not require individual finger movement to grasp objects. Instead, it utilises sensors embedded in its ‘skin’ to predict if it will drop objects, making it an energy-efficient and easier-to-control alternative to fully motorized robotic hands.
The Cambridge team’s innovative design holds potential for the creation of low-cost robots capable of more natural, human-like movement. The robots could also learn to handle a diverse array of objects, significantly expanding their application range. Their research findings have been published in the journal Advanced Intelligent Systems.
In nature, movement is a harmonious interplay between the brain and the body, allowing humans and animals to execute complex motions without expending excessive energy. Over recent years, the advent of advanced 3D printing techniques has enabled the integration of soft components into robotics design. This has opened up new avenues for creating simple, energy-efficient systems with added complexity.
At the Bio-Inspired Robotics Laboratory in Cambridge’s Department of Engineering, under the guidance of Professor Fumiya Iida, researchers are exploring solutions to this challenge. Their focus is on developing a robotic hand capable of grasping various objects while applying the correct amount of pressure and using minimal energy.
Dr Thomas George-Thuruthel, co-author of the study and currently based at University College London (UCL) East, explained their approach. He said, “Our lab has previously demonstrated that significant motion range in a robot hand can be achieved merely by moving the wrist. We wanted to investigate if a robot hand employing passive movement could not only grasp objects but also predict and adapt if it was likely to drop them.”
The team used a 3D-printed anthropomorphic hand fitted with tactile sensors to allow the hand to sense what it was touching. The hand was designed for passive, wrist-based movement. Dr Kieran Gilday, the study’s first author who is now based at EPFL in Lausanne, Switzerland, explained, “This type of hand has a spring-like quality. It can pick things up without any finger actuation. The tactile sensors provide the robot with feedback on how well it is gripping an object, so it knows when it’s starting to slip. This allows it to predict potential failures.”
The robot learned successful gripping techniques through trial and error. After training with balls, it attempted to grasp various objects such as a peach, a computer mouse, and a roll of bubble wrap. In these tests, the hand successfully grasped 11 out of 14 objects.
Professor Iida emphasized the benefits of this design. He said, “The big advantage is the range of motion we can achieve without using any actuators. We aim to simplify the hand as much as possible. We can obtain a lot of valuable information and a high degree of control without any actuators. When we do add them, we’ll achieve more complex behaviour in a more efficient package.”
A fully actuated robotic hand presents complex control problems besides requiring significant energy. The passive design of the Cambridge-created hand, with its limited number of sensors, is easier to control and offers a broad range of motion. It also simplifies the learning process.
For those interested in this innovation in electronics and robotics or those involved in programming languages or coding for similar projects, this development presents exciting possibilities for future advancements in robotics.
For more information about this breakthrough in robotics technology, contact Sarah Collins at +012-237-65542 or via email at [email protected]