Production Made More Effective with Next-Gen Robotic Hands

The HAND research center, backed by the National Science Foundation, is tackling one of robotics’ greatest challenges: giving robot hands the fluid motion and fine motor control of human hands. By combining soft, compliant materials with advanced sensors and artificial muscle technologies, HAND is developing robotic end-effectors that can move, grasp, and adjust with unprecedented precision. These hands won’t just mimic motion—they’ll sense texture, stiffness, and even temperature, allowing for real-time feedback and adaptive control. With experts from robotics, softbotics, and human-computer interaction, and led by Carnegie Mellon, HAND is redefining how robots move—and how they move us into the future of intelligent manufacturing.

Transcript

00:00:08 HAND is an engineering research center that's  sponsored by the National Science Foundation.   Its objective is to create the next generation  of robot hands. These next generation hands   will have the same dexterity and  mobility as natural human hands.   This represents a grand challenge in robotics and  then also the future of manufacturing. With these   robot hands, what we like to do is  create robot companions that these   skilled laborers and this workforce  can work side by side with to make   production much more effective so that these  small manufacturing enterprises can compete   with the much larger multinational global  manufacturers. To achieve these goals, we   have a large partnership across several different  universities and also industrial partnerships. So a lot of the existing robot end-effectors  are engineered using rigid materials. This  

00:01:02 is completely fine for more simple  pick-and-place operations. Oftentimes,   these robot systems rely on cameras and  computer vision in order to provide feedback   and to compensate for any kind of errors in those  physical interactions. However, when we think   about using these robot hands at a much wider  range of applications, we have to design them so   they become much more robust. We have to design  them with materials that are soft, compliant   so they can conform to the surfaces that  these robot end-effectors come in contact with. These next-generation robot hands need to have  a sense of touch. They should be able to to   feel objects, detect physical properties like  the stiffness of the material, its fragility,   its surface texture, in some cases, even its  temperature or how wet or lubricated the surface   is. If we want to achieve hands that have the  same fluidity and dexterity as natural human  

00:02:00 hands and can also be mounted to a wider range  of robot arms, we have to replace a lot of these   motors with next-generation artificial muscle  technologies. And the goal with these artificial   muscle technologies is to replace motors with  actuators that mimic a lot of the same properties   as natural skeletal muscle. So they would be  much more lightweight, potentially 3-D printable,   and they can be integrated with sensors  and then other elements of the robot hand. The researchers involved in HAND come from a  variety of different scientific disciplines   and backgrounds. And I truly believe  that Carnegie Mellon is one of the best   places to lead this type of research  given its strengths in softbotics,   robotics, human computer interaction, also  advancements in materials and electronics.