DERMIS: End-to-End Design of a Fully Integrated Large-Area Grasp-State-Adaptive Tactile Sensor System on a-IGZO TFT

Area of Science:

• Robotics and Artificial Intelligence
• Materials Science and Engineering
• Sensor Technology

Background:

• Existing tactile sensors often lack comprehensive grasp state analysis.
• Current systems struggle with independent shear and normal force detection.
• Large-area electronic skins require integrated, low-power sensing solutions.

Purpose of the Study:

• To design and implement a high-resolution, fully-integrated tactile sensor system (DERMIS) for large-area electronic skins.
• To develop a sensor capable of measuring multiple contact cues, including friction and forces, for adaptive grasping.
• To demonstrate an end-to-end design strategy from sensor to on-chip feature extraction.

Main Methods:

• Utilized flexible thin-film transistor (TFT) technology for large-area implementation.
• Developed a novel differential capacitive sensor structure for independent shear and normal force sensing.
• Implemented an analog front-end for direct, on-chip extraction of contact parameters, avoiding complex algorithms.

Main Results:

• Achieved a human-like force resolution of 2 mN_RMS at a 0.6 mm pitch.
• Demonstrated a low per-taxel power consumption of 72 µW and area of 0.36 mm².
• Successfully prototyped the system on a 9x4 mm² large-area substrate, measuring friction, contact, and lift-off cues.

Conclusions:

• The DERMIS system offers a biologically-inspired, grasp-state-adaptive tactile sensing solution.
• Analog-based encoding of contact parameters simplifies processing and enhances efficiency.
• This work paves the way for advanced robotic manipulation and human-computer interaction through sophisticated electronic skins.