Thin and soft optical tactile sensor for highly sensitive object perception

Area of Science:

• Robotics and Mechanical Engineering
• Optical Sensing Technologies
• Materials Science

Background:

• Tactile sensing is crucial for safe human-robot and human-wearable interaction.
• Existing optical tactile sensors often rely on complex, bulky optical assemblies.
• There is a need for compact, compliant, and alignment-free tactile sensing solutions.

Purpose of the Study:

• To develop a thin, compact, and compliant optical tactile sensor with an alignment-free architecture.
• To demonstrate the sensor's capability for high-resolution tactile information acquisition.
• To validate the sensor's multimodal sensing and robustness to environmental factors.

Main Methods:

• Utilizing deformation-induced speckle pattern variations within a soft silicone medium.
• Employing a data-driven model for inferring tactile information from speckle patterns.
• Integrating the sensor into a robotic gripper for practical application testing.

Main Results:

• Achieved a spatial resolution better than 20 µm and force measurement precision below 40 mN.
• Demonstrated simultaneous estimation of contact position, applied force, and temperature.
• Successfully classified nine engraved surface patterns with 93.33% accuracy using a robotic gripper.

Conclusions:

• The proposed speckle-based optical tactile sensor offers a compact, manufacturable, and compliant solution.
• This technology is a promising advancement for next-generation soft robotics and wearable haptic systems.
• The alignment-free design and multimodal sensing capabilities address key limitations of current optical tactile sensors.