Magnetic Tactile Sensor with Bionic Hair Array for Sliding Sensing and Object Recognition

Abstract

Due to the high application value in intelligent robots, tactile sensors with large sensing area and multi-dimensional sensing ability have attracted the attention of researchers in recent years. Inspired by bionics of hairs on human skin, a flexible tactile sensor based on magnetic cilia array is developed, showing extremely high sensitivity and stability. The upper layers of the sensor are multiple magnetic cilia containing magnetic particles, while the lower layer is a serpentine flexible circuit board with a magnetic sensor array. When magnetic cilia are bent under force, the magnetic sensor array can detect changes in the magnetic field, thereby the magnitude and direction of external force can be obtained. The proposed sensor has a resolution of 0.2 mN with a working range of 0-19.5 mN and can distinguish the direction of external force. The large sensing area and short response time make this sensor suitable for sliding tactile detection, and experiments show that the sensor can be also applied in object recognition with a success accuracy of 97%. In addition to the shape of objects, the sensor can identify whether there is magnetism inside objects, making it of significant value in intelligent robots and modern medicine.

Keywords: flexible electronics; magnetic material; object recognition; sliding sensing; tactile sensor.

Figure 1

Hairs on human skin, structural diagram of our sensors, and sensor applications. a) Hairs on the human skin and the nerves under the skin. b) Schematic and exploded view of our sensor, with the right image showing the magnetic field distribution when a single cilium is bent. c) Schematic diagram of object recognition with our sensor. After the robotic arm with our sensor grabs an object, the recognition results and whether the object is magnetic can be displayed on a screen.

Figure 2

Simulation of our sensor. a) Simulation of stress and displacement when magnetic cilia are bent. b) Simulation of magnetic distribution of cilia array.

Figure 3

Production process and circuit design of our sensor. a) Production process of our sensor. b) VSM test results of mixtures with different magnetic particle contents. c) Unit volume remanence and Young's modulus of mixtures with different magnetic particle contents. d) Circuit design of our sensor.

Figure 4

Microscopic images of mixtures with different magnetic particle contents and photos of cilia arrays made by demolding method. a) Microscopic images of mixtures with different magnetic particle contents. When the content exceeds 50%, magnetic particles aggregate and bubbles appear in mixtures. b) Single row magnetic cilia with different heights. c) Magnetic cilia arrays with different heights.

Figure 5

Performance results of the sensor. a) Response of the sensor to force. Inset: sensor noise without input. b) Response of the X‐axis and Y‐axis of the magnetic sensor to eight different directions of force. The eight red arrows below the curve indicate the force directions. c) Sensor output when the cilium is repeatedly bent under low, medium, and high pressure. Inset: output waveform of three bends. d) Test results of dynamic response time of our sensor. e) Consistency in the output of six sensors manufactured using the same process. f) The hysteresis of the magnetic sensor MLX90393.

Figure 6

Application of our sensor on a robot flexible gripper. a) The left image shows eight magnetic sensors on a serpentine FPC board. The middle image is a photo of the robot flexible gripper. The right image shows a LabVIEW software, which can display the position and size of external force in real‐time. The darker the color, the greater the force. b) The output changes over time of eight sensors when a finger slides across the cilia array. c) Photo of object recognition with our flexible gripper. When an object is picked up by the gripper, the screen will display recognized results and whether it is magnetic. d) Confusion matrix of recognition results for eight different objects.