Open Collections

Wani, Sadan

University of British Columbia

The proliferation of augmented reality in professional settings has led to innovation in Electronic skin (E-skin) sensing technologies applied to human controlled robots. E-skin can enable machines to perceive the world in a similar light to human touch, with applications in medical robotics and smart prosthetics. This thesis discusses one such multimodal capacitive elastomeric skin, capable of shear, pressure, and proximity sensing. Object detection through proximity sensing may benefit a robot by reducing the need for complex visual processing to operate tools in real world settings and allow for crash/impact detection if vision is obstructed. The author presents developments in three areas, namely: Advancements of fabrication techniques of e-skin, Soft capacitive proximity sensor (SCPS) electrode designs, and low cost open-source proximity characterization setups. The innovations in fabrication led to better interfacing between the soft and rigid electrical components of the sensing system, improving reliability and durability against delamination forces of up to 50 N. Investigations into proximity electrode designs through simulation and experimental testing, resulted in object detection ranges of 30 mm when the sensor was placed flat, and over 25 mm when the sensor was curved onto a digit of radius 8.8 mm. Testing of multiple side by side proximity sensors showed negligible ( < 1%) performance impact. The difference in capacitance readings for a grounded vs ungrounded conductor was documented and a rudimentary theory to explain the discrepancy is proposed. A 3-Axis characterization setup on which the quantitative experimentation was conducted is devised to be inexpensive, with inductive sensing to allow for contactless “homing” to the top of the soft multimodal sensors. It is shown that the device is capable of up to 50 m of displacement accuracy and high repeatability, while being “invisible” to the sensors being characterized. The setup enabled detailed characterization of the SCPSs in the Z axis and preliminary characterization in the X and Y axes. Future work is intended to explore mapping of SCPSs 3D spatial performance, machine learning to devise feedback controls, mounting on humanoid robotic hands, classification of materials and object shape detection.

Text; Moving Image; Dataset

2025-04-30

Attribution-NonCommercial-NoDerivatives 4.0 International

http://hdl.handle.net/2429/88145

Biomedical Engineering

University of British Columbia

UBCV

http://creativecommons.org/licenses/by-nc-nd/4.0/