ERF Magnetic Tactile Sensors for Soft Robotic Hand

What it does

The ERF (Electro-Repulsive Force) magnetic tactile sensor solves the problem of enabling soft robotic hand to feel and react to objects. It provides crucial touch feedback and passively dampens motion, maintaining the robotic hand's natural compliance.

Your inspiration

Soft robotic hands have been developed significantly over the past few decades, offering a solution to the limitations of traditional rigid robotic systems. By using soft and flexible materials, the robotic hand can grasp objects while minimizing the risk of damage to the object itself. However, it faces several challenges, particularly concerning material durability. Silicone, a commonly used material, tends to degrade over time due to repeated bending and rotational movements. Therefore, we propose a solution that addresses the issue of material endurance while preserving the flexible nature of the robotic hand.

How it works

The ERF (Electro-Repulsive Force) tactile sensor is constructed with two N35 neodymium magnets, each 13mm in diameter, and a coil formed from 80 turns of 0.15mm soft copper wire. When external forces, such as touch or object contact, induce displacement in one of the magnets, the resultant relative motion between the magnets and the coil generates a voltage. This phenomenon is precisely described by Faraday's Law of Induction. The induced voltage is subsequently filtered, amplified, and then transmitted to a microcontroller, where it functions as a tactile signal for a tendon wire-driven soft robotic hand. These innovative sensors are strategically embedded within each finger segment of the robotic hand. A key feature of the ERF sensor is the inherent repulsive force between its magnets, typically ranging from 1 to 45 N. This force serves as a passive damping mechanism, crucial for preserving the soft robotic hand's compliant motion and flexible nature.

Design process

Initially, we encountered challenges in the manufacturing process of the sensor casing using a 3D printer. One of the critical considerations was the selection of suitable magnets and copper wire. The magnets needed to possess sufficient repulsive force to function effectively as part of a tactile sensing mechanism, yet not be excessively strong, as the 3D-printed casing has limited structural strength to withstand opposing magnetic poles positioned at a minimum distance of 10 mm. After evaluating various options, we selected N35-grade neodymium magnets as an appropriate compromise between force and mechanical compatibility. Additionally, we chose soft copper wire with a diameter of 0.15 mm. The smaller wire diameter allows for a higher number of coil windings, which increases the sensor’s output voltage and improves resistance to electrical noise. Under typical operating conditions, the sensor generates an output of approximately 300 mV, with noise levels ranging between 5–10 mV. To amplify the signal, we implemented a non-inverting operational amplifier circuit, along with a low-pass filter (LPF) featuring a cutoff frequency of 10 Hz, to effectively attenuate high-frequency noise components. Currently, the ERF tactile sensor can be read properly by ESP32 microcontroller.

How it is different

Unlike traditional soft robotic hands that rely on silicone and tactile sensors, this design eliminates direct contact between moving materials by using magnetic repulsion to sense touch. This reduces mechanical wear and significantly increases durability. In addition, most soft hands lack integrated sensing or use bulky sensors that affect flexibility. Our ERF sensor is compact, embedded, and uses passive sensing through induction, requiring no external excitation. This approach combines durability, compactness, and functionality without compromising hand mobility. The result is a soft robotic hand that can grasp objects gently, sense contact, and last longer under repeated use.

Future plans

We aim to integrate soft robotic hand to the robotic hand with a robotic glove system to enable bidirectional interaction. The robotic glove captures the human hand's movements and sends positional data to control the soft robotic hand in real time. At the same time, tactile feedback from ERF (Electro-Repulsive Force) magnetic sensors embedded in the soft robotic hand is transmitted back to the robotic glove, allowing the user to feel touch and pressure. This setup enhances teleoperation robotic control allow precision and brings more natural interaction between humans and robotic systems.

Awards