The dexterity of a human hand has inspired a team of researchers from the National University of Singapore (NUS) to develop a reconfigurable hybrid robotics system capable of gripping a variety of items, ranging from small, soft, and delicate to enormous, heavy, and monstrous. This technology is expected to have an impact on a range of industries, including food processing, vertical farming, and fast-moving consumer goods packaging, all of which will be able to automate more of their processes in the coming years.
Soft, adaptable 3D-printed fingers with a customizable gripper base are used in the hybrid robotic grippers. Under the staff’s start-up RoPlus (RO+), which includes NUS researchers, the robotic innovation is now in the process of being marketed to commercial partners.
The shape, texture, weight, and size of an object influence how we hold it. One of the key reasons why many companies still rely largely on human labor to package and handle delicate commodities is because of this. Our innovative robotic gripper technology revolutionizes standard pick-and-place activities by allowing robots to securely interact with fragile things of diverse shapes, sizes, and stiffness, just like a human hand.” — Raye Yeow, Associate Professor, National University of Singapore’s Department of Biomedical Engineering and Advanced Robotics Centre
Gripping is one of the most common and natural human activities, however, it isn’t as natural for robots. To achieve human-like gripping abilities, robots use computer vision and deep learning to determine the type of objects in front of them as well as their orientation. The gripper can then determine how best to choose and put objects mechanically, reducing the need for extensive human intervention.
The NUS team devised hybrid robotic grippers, which are made up of three or four soft fingers that can reconfigure on demand, in order to develop robotic grippers that are as dexterous as human fingers. The fingers are powered by air and include a unique locking mechanism for adjusting stiffness. The NUS team has created three types of hybrid robotic gripper methods—virtually three separate fingers—that could be useful in a variety of situations.
The first is GourmetGrip, which is designed for the most granular tasks, such as handling delicate bite-sized snacks or meals that are simply prone to harm, such as tofu, and puts them into take-out bins. This soft-handed mode can be reconfigured to meet a wide range of grip stances as well as a variety of space constraints.
The second type of gripper is the UnisoGrip, or Universal Soft Gripper, which is the team’s most widely applicable solution. It’s made for dealing with packaged things as they’re being placed into bins for distribution and transportation beside the assembly line. It contains soft rotating gripper fingers for gentle hold, as well as a vacuum suction cup that allows it to transfer more awkwardly positioned objects, such as the nook of a tote bin.
The third type of gripper, based on the GourmetGrip/UnisoGrip platforms, is completely configurable to meet specific customer needs and space limits. This approach provides a wide range of grasping options for objects of varied shapes, sizes, and packing materials.
NUS has also issued recommendations on how society and organizations should approach Artificial Intelligence (AI) in ways that actually promote human interests and well-being. The manifesto “The Road to a Human-Centred Digital Society: Opportunities, Challenges, and Responsibilities for Humans in the Age of Machines” promotes a strategy that prioritizes human feelings of competence, belonging, control, and well-being. It includes seven high-level recommendations to help businesses and policymakers pursue a Human-Centered AI approach (HCAI).