Robots are becoming increasingly capable of performing everyday tasks, but many still struggle with delicate movements that humans take for granted.
Picking up fragile items such as fruit, glasses, or potato chips without damaging them requires a very precise sense of touch.
Now, researchers at the University of Texas at Austin have developed a robotic hand that can do exactly that.
The new technology allows robots to gently grasp extremely fragile objects, including raspberries and potato chips, without crushing them.
The system is called Fragile Object Grasping with Tactile Sensing, or FORTE. It combines soft robotic fingers with advanced sensing technology that gives the robot a much better sense of touch.
The research was led by Siqi Shang, a doctoral student in electrical and computer engineering at UT Austin. Shang explained that while robots have improved at performing large actions, such as folding clothes or moving objects around a room, they often struggle with smaller, more delicate tasks.
For example, a robot might be able to fold a shirt, but picking up a pair of glasses or carefully unpacking fruit from grocery bags can still be very difficult.
The researchers believe that giving robots better tactile feedback—similar to the way human hands sense pressure—will help them perform these kinds of tasks more safely.
The robotic fingers used in this system were inspired by a natural design known as the fin-ray effect, which comes from the structure of fish fins.
Fish fins can bend and adjust easily to different shapes, allowing them to interact gently with their surroundings. Using this idea, the researchers created flexible robotic fingers using advanced 3D printing techniques.
Inside each finger are small air-filled channels that act as touch sensors. When the robotic hand begins to grasp an object, the shape of the fingers changes slightly. This movement alters the air pressure inside the channels. Small sensors detect these pressure changes and send the information to the robot in real time.
By analyzing these signals, the robot can tell how much force it is applying to the object. It can also detect when the object is starting to slip, allowing it to quickly adjust its grip before the object falls.
The research team tested the system on 31 different objects. These included fragile items such as raspberries and potato chips, slippery objects like jam jars and billiard balls, and common household items such as apples and soup cans. The robotic hand successfully grasped objects in nearly 92 percent of the tests, which is better than many traditional robotic grippers that rely mainly on cameras.
The system was also very good at detecting when objects began to slip. It correctly identified 93 percent of slipping events and did so with perfect precision, meaning it never falsely signaled a slip.
According to the researchers, the sensors respond quickly and operate on timescales similar to those of human touch. This allows the robot to apply what scientists call the “Goldilocks” level of pressure—not too strong and not too weak.
The technology could have important uses in several industries. Robots with sensitive touch could help food companies handle fragile products, assist doctors in medical laboratories, or manage delicate electronic components during manufacturing.
The researchers have made the designs and software publicly available so that other scientists can build on their work and further improve robotic touch.
