Imagine a robot that can feel a touch and instantly react without using a computer, batteries, or electronic sensors. It may sound like science fiction, but researchers have now made it possible.
A team of scientists from the National University of Singapore (NUS) has developed a groundbreaking soft force sensor that allows soft robots to sense pressure and respond immediately using only mechanical movement and fluid flow. The new technology removes the need for electronic sensors, processors, and powered control systems, making robots simpler, more reliable, and better suited for harsh environments.
The research, led by Professor Benjamin Tee and Professor Cecilia Laschi, was published in the journal Science Advances and could transform fields ranging from healthcare and prosthetics to underwater exploration and disaster response.
Why Soft Robots Need a Better Way to Sense
Unlike traditional robots made of hard metal parts, soft robots are built from flexible, rubber-like materials. Their soft bodies make them much safer for interacting with people and delicate objects.
Because of this flexibility, soft robots are considered ideal for tasks such as:
Minimally invasive surgery
Assisting elderly people
Handling fragile objects
Exploring deep oceans
Working in dangerous environments
However, these robots have faced one major challenge.
To detect touch or pressure, most soft robots still rely on electronic sensors connected to processors and powered actuators. When the robot feels something, the sensor sends electrical signals to a computer, which processes the information and tells the robot what action to take.
This system works well under normal conditions, but it has several drawbacks.
Electronics add extra weight, increase manufacturing complexity, consume power, and can fail in environments with water, high temperatures, strong pressure, or electromagnetic interference.
The NUS researchers wanted to remove this entire electronic chain.
A Sensor That Works Like Nature
The new invention is called ME-SOFS, short for Mechanical Soft Force Sensor.
Instead of converting touch into electrical signals first, the sensor converts physical force directly into fluid movement. That flowing fluid immediately activates the robot's movement.
In simple words:
Touch → Fluid Flow → Robot Movement
There is no computer, processor, battery-powered electronics, or software involved in this process.
This creates an almost instinct-like reaction, similar to how living organisms often respond quickly to their surroundings.
How Does ME-SOFS Work?
The sensor is produced using 3D printing and is made from a soft porous material.
At its center sits a flexible pillar connected to five tiny chambers filled with fluid.
Four chambers detect forces from different horizontal directions.
One chamber detects vertical pressure.
When pressure is applied, the central pillar bends toward the force.
As it bends, it squeezes one or more fluid chambers.
The compressed fluid then flows through soft tubes to connected actuators, causing the robot to move immediately.
Each chamber works independently, allowing the system to identify exactly where the force came from.
The sensor can detect forces in three dimensions:
Left and right
Forward and backward
Up and down
Everything happens mechanically, without requiring electronic signal processing.
Producing Electrical Signals Without Electronics
Although the robot can operate completely mechanically, the researchers also created a clever way to measure force electronically without powered sensors.
As the fluid moves, it pushes tiny magnets past specially designed metal arcs.
This changing magnetic field naturally generates small voltage pulses through electromagnetic induction—the same physical principle used in a bicycle dynamo that produces electricity while the wheels spin.
The number of pulses corresponds directly to the amount of force applied.
This allows researchers to record and analyze forces without adding traditional electronic sensing systems.
Sensitivity Can Be Customized
One of the biggest advantages of ME-SOFS is its flexibility.
Its sensitivity depends mainly on the shape of its internal structure, including:
Hole size
Wall thickness
Angle of the central foam structure
By simply modifying the 3D printing design, engineers can create sensors that are more or less sensitive depending on the application.
This makes the technology highly adaptable for different robotic systems.
Inspired by Nature
Professor Benjamin Tee explained that the team took inspiration from biological systems.
Many living organisms rely on fluid-filled cells to sense touch and transmit mechanical information.
The researchers wondered whether fluid channels alone could both detect touch and directly provide tactile feedback.
Their experiments showed that this fluid-based design is extremely robust and could become an important building block for future physical AI systems.
A Soft Glove That Can Feel Weight
To demonstrate the technology, the team built a soft robotic glove.
Each fingertip contained a miniature ME-SOFS sensor about the size of a green pea.
Remarkably, the entire glove was 3D printed in one continuous process using a single material, requiring no manual assembly.
When someone wore the glove and picked up objects, the sensors measured the pressure at each fingertip.
Using this information, the system could estimate the weight of the object being held.
Such technology could improve:
Prosthetic hands
Human-machine interaction
Virtual reality systems
Industrial robotics
Feeling Through a Robot
The researchers also created an impressive touch-feedback system.
A robotic arm equipped with ME-SOFS sensors was connected through fluid channels to a soft pad worn on a person's fingertips.
The human operator wore a blindfold and controlled the robotic arm using only touch.
As the robotic gripper squeezed different objects—including a fragile egg, wooden blocks, and a half-filled water bottle—the fluid pressure traveled directly back to the person's fingertips.
Without seeing anything, the operator could feel exactly how hard the robot was gripping.
This demonstrates true mechanical touch communication without electronic processing.
Teaching Robots by Touch
The researchers also recorded the successful gripping forces used during these experiments.
Later, they replayed the same force patterns so the robot could automatically repeat the correct grasping motion.
This shows that ME-SOFS can both provide real-time tactile feedback and collect valuable data for robot training.
Future robots could learn safe and effective movements simply by replaying successful force patterns.
More Than Just Robotics
The technology also proved useful in several other applications.
The researchers used it to guide tiny liquid droplets through miniature fluid control systems without using software.
This could become valuable for portable medical diagnostic devices.
They also connected the sensor to flexible hair-like structures that bent automatically depending on the direction and strength of applied force.
These demonstrations show how the same technology can be adapted far beyond traditional robotics.
Built for Extreme Environments
One of ME-SOFS's greatest strengths is its durability.
The sensor continued working reliably in:
Hot water at 90°C
High-pressure conditions equal to approximately 11 meters underwater
Its open fluid channels naturally balance with surrounding water pressure.
As a result, the sensor measures only the applied force instead of being affected by environmental pressure.
Since it contains no electronic components, it is also immune to electromagnetic interference that can disrupt conventional electronic sensors.
These qualities make it especially attractive for underwater exploration, industrial inspection, and other harsh environments.
A New Kind of Mechanical Intelligence
Professor Cecilia Laschi describes this technology as an example of embodied intelligence.
Instead of relying on a computer brain to process every signal, the robot's body itself performs part of the intelligence through its physical design.
Nature often uses this strategy.
Many animals and plants respond to their environment because of how their bodies are built, not only because of complex nervous systems.
ME-SOFS brings a similar concept into robotics, allowing machines to react almost instinctively.
Looking Ahead
The researchers believe this technology has enormous future potential.
Because it can be miniaturized and scaled for different force levels, ME-SOFS could be integrated into many kinds of soft robots.
Possible future applications include:
Medical training systems where students can feel exactly what an expert feels during surgery.
Elderly care robots that instantly provide extra support if a person begins to fall.
Advanced prosthetic limbs with natural touch feedback.
Deep-sea exploration robots that operate where electronics often fail.
Search-and-rescue robots working in dangerous environments.
By replacing complex electronic systems with direct mechanical sensing and action, ME-SOFS represents a major step toward safer, smarter, and more reliable soft robots. As robotic technology continues to evolve, this innovation could help create machines that respond to the world almost as naturally and instinctively as living organisms.
Reference:
- Zhexin Xie et al.
Comments
Post a Comment