In automated sorting and precision assembly, certain types of workpieces have always been a headache: freshly baked pastries, highgloss mobile phone glass covers, thin-walled ceramic parts, fresh strawberries – some are too soft and easily deformed, others too brittle and prone to breakage, and many have extremely scratch-sensitive surfaces. Conventional grippers either fail to hold them securely or damage them upon grasping. Today, with the integration of precision force-controlled grippers and flexible electronic skin, a truly “dexterous hand” is moving from the laboratory into industrial and everyday applications, giving machines a grasping intelligence that approaches that of humans.

Part 1: The Precision Force-controlled Gripper – A Mechanical Hand That “Weighs” Its Force

What is a precision force-controlled gripper? In simple terms, it is an intelligent gripping device capable of sensing and actively adjusting gripping force in real time.

Core Working Principle

Conventional grippers typically use position control – the fingers move to a fixed position regardless of whether they are grasping an iron block or a piece of tofu. This is clearly unreasonable. A precision force-controlled gripper, on the other hand, integrates highsensitivity force sensors (such as strain-gauge or piezoelectric force sensors) together with a real-time closed-loop feedback control system to form a force control loop:

Sensing – The force sensor detects the contact force between the gripper fingers and the workpiece at millisecond intervals.

Evaluation – The controller compares the measured force with a preset target force.

Adjustment – The drive motor (or voice coil actuator, piezoelectric actuator) dynamically adjusts the gripping force to keep it within the optimal range.

This process repeats thousands of times per second, so regardless of how the workpiece dimensions vary within tolerance, the force applied by the gripper remains constant and precise.

Key Capability: Adapting to Different Material Properties

Different materials have vastly different properties (stiffness, elastic modulus, fragility). A precision force-controlled gripper can be preprogrammed or use adaptive algorithms to match the appropriate gripping force for each workpiece type:

Rigid metal parts – Higher gripping force can be used to ensure no slipping during positioning.

Elastic rubber parts – Force control prevents excessive compression that would cause deformation.

Fragile ceramics or glass – The gripping force is precisely kept below the fracture threshold, while a friction model ensures the part does not slip.

Soft foods (cakes, bread) – Force control achieves a “gentle embrace” grasp, causing only minimal indentation without destroying the structure.

It is this ability to “weigh” the force that enables precision force-controlled grippers to handle precision manufacturing and delicate object sorting tasks that conventional grippers cannot manage.

Part 2: Flexible Skin – Giving the Gripper Tactile Perception

Even with precise force closed-loop control, a gripper still lacks a crucial capability: spatially resolved touch. It knows how much total force it applies, but it does not know exactly where that force is distributed across the finger surfaces, nor whether the contact surface is smooth or rough, nor whether the object is starting to slip.

To solve this problem, scientists wrap flexible electronic skin around the gripper fingers. As described in a previous article, this skin contains a flexible piezoelectric sensor array that, like human fingerprints and tactile corpuscles, converts the contact pressure distribution into electrical signals in real time.

New Capabilities Brought by Flexible Skin

Contact localization – The sensor array on the skin can precisely determine where the object contacts the gripper fingers, allowing the control system to optimise the grasping posture and avoid stress concentration from point contacts.

Slip detection – When the object begins to move slightly between the fingers, the skin detects changes in the pressure distribution, and the control system immediately adjusts the gripping force or finger posture to prevent dropping.

Surface texture recognition – By analysing microvibration signals on the contact surface, the dexterous hand can distinguish submillimetre surface patterns and even identify the material (e.g., paper versus fabric).

Multimodal sensing – In addition to piezoelectric sensors, temperature and humidity sensors can be integrated, allowing the gripper to perceive whether an object is hot or cold, dry or wet.

Part 3: Birth of the Dexterous Hand – Integration of Force Control and Tactile Sensing

When the “muscle strength” of the precision force-controlled gripper is combined with the “neural sensing” of flexible skin, the basic architecture of an advanced dexterous hand is established.

Such a dexterous hand demonstrates unprecedented advantages when handling complex tasks.

Example 1: Grasping a Ripe Strawberry

Conventional gripper – Either fails to hold it or crushes it.

Dexterous hand – The flexible skin first lightly touches the strawberry, sensing its contour and local stiffness distribution. The precision forcecontrolled gripper closes with a very low force; as soon as it detects slight deformation of the strawberry surface (feedback from the skin array), it locks the current force value. After grasping, it continuously monitors slip trends and makes fine adjustments as needed, safely transferring the strawberry to an inspection station.

Example 2: Assembling a Precision Optical Lens Barrel

Conventional method – Requires dedicated fixtures and manual assistance.

Dexterous hand – The skin senses the relative position deviation between the lens barrel and the housing. The precision force control adjusts the grasping posture and completes the insertion in a compliant manner, avoiding scratches or jamming.

Part 4: From Industrial Dexterous Hands to Biomimetic Dexterous Hands

Precision force-controlled grippers equipped with flexible skin are already being used in intelligent sorting, precision assembly, and food processing. But this is only the beginning. With advances in flexible electronics, micro-actuators, and edge computing, future dexterous hands will approach the performance of the human hand even more closely:

Multi-finger cooperation – Not just two-finger grippers, but five independently force-controlled fingers, each covered with flexible skin, capable of pinching, holding, twisting, and other fine manipulations.

Autonomous learning – Combined with artificial intelligence, a dexterous hand can learn optimal grasping strategies for unknown objects through repeated trial and error.

Tactile feedback – Tactile signals perceived by the dexterous hand can be transmitted back to a human operator (e.g., through a vibrating glove), enabling “tele-presence of touch” for delicate operations in hazardous environments.

Conclusion

The precision force-controlled gripper solves the problem of “how much force to use”; flexible skin solves the problems of “where the force is, whether it is slipping, and what the surface is like.” Together, they give the machine a true sense of “hand feel.” This dexterous hand not only makes automated equipment safer, more reliable, and more intelligent, but also paves the way for cuttingedge fields such as humanoid robots, intelligent prosthetics, and remote surgery. The next time you see a robot effortlessly pick a rose without damaging its petals, remember: that is the “precision dance” performed by precision force control technology and flexible tactile skin working in harmony.