Prof. Kenneth J Loh

Machine vision-based structural health monitoring is gaining popularity due to the rich information one can extract from video and images. However, the extraction of characteristic parameters from images often requires manual intervention, thereby limiting its scalability and effectiveness. In contrast, deep learning overcomes the aforementioned shortcoming in that it can autonomously extract feature parameters (e.g. structural damage) from image datasets. Therefore, this study aims to validate the use of machine vision and deep learning for structural health monitoring by focusing on a particular application of detecting bolt loosening. First, a dataset that contains 300 images was collected. The dataset includes two bolt states, namely, tight and loosened. Second, a faster region-based convolutional neural network was trained and evaluated. The test results showed that the average precision of bolt damage detection is 0.9503. Thereafter, bolts were loosened to various screw heights, and images obtained from different angles, lighting conditions, and vibration conditions were identified separately. The trained model was then employed to validate that bolt loosening could be detected with sufficient accuracy using various types of images. Finally, the trained model was connected with a webcam to realize real-time bolt loosening damage monitoring.

17 October 2007 in Journal of Intelligent Material Systems and Structures

Journal Article

Tailoring Piezoresistive Sensitivity of Multilayer Carbon Nanotube Composite Strain Sensors

Kenneth J Loh

J.P. Lynch

B.S. Shim

N.A. Kotov

In recent years, carbon nanotubes have been utilized for a variety of applications, including nanoelectronics and various types of sensors. In particular, researchers have sought to take advantage of the superior electrical properties of carbon nanotubes for fabricating novel strain sensors. This article presents a single-walled carbon nanotube (SWNT)-polyelectrolyte (PE) composite thin film strain sensor fabricated with a layer-by-layer (LbL) process. Optimization of bulk SWNT-PE strain sensor properties is achieved by varying various LbL fabrication parameters, followed by characterization of strain-sensing electromechanical responses. A resistor and capacitor (RC)-circuit model is proposed and validated with electrical impedance spectroscopy to fit experimental results and to identify equivalent circuit element parameters sensitive to strain. Experimental results suggest consistent trends between SWNT and PE concentrations to strain sensor sensitivities. Simply by adjusting the weight fraction of SWNT solutions and film thickness, strain sensitivities between 0.1 and 1.8 have been achieved. While SWNT-PE strain sensitivity is lower than some metal-foil strain gauges ($2), the LbL method allows for precise tailoring of the properties (i.e., strain sensitivity, resistivity, among others) of a high-capacity (±10,000 μm m^-1) homogeneous multilayer strain sensor.

About Me

Dr. Ken Loh is the TaylorMade Golf Chancellor’s Endowed Professor in the Department of Structural Engineering at UC San Diego and previously served as the Department Vice Chair (2018-2021). He is the Director of the Active, Responsive, Multifunctional, and Ordered-materials Research (ARMOR) Lab and is the Director of the Jacobs School of Engineering, Center for Extreme Events Research (CEER). He is also an affiliate faculty member of the Materials Science & Engineering Program. Dr. Loh received his B.S. in Civil Engineering from Johns Hopkins University in 2004. His graduate studies were at the University of Michigan, where he completed two M.S. degrees in Structural Engineering (2005) and Materials Science & Engineering (2008), as well as a Ph.D. in Structural Engineering in 2008. He started his Assistant Professor career in January 2009 in the Department of Civil & Environmental Engineering at UC Davis, before joining UC San Diego in January 2016. His research interests are in multifunctional and stimuli-responsive materials, tomographic imaging techniques, wearable sensors, active metamaterials, and soft material actuators applied towards solving problems related to human performance, structural sustainment, and human-structure interactions. In addition to his academic career, Dr. Loh is an Engineering Duty Officer in the U.S. Navy Reserve and a co-founder of a start-up company, JAK Labs, Inc.

Prof. Kenneth J Loh

Prof. Kenneth J Loh

Research Focus & Metrics

Academic Career Length

Publication Type Breakdown

Representative Publications

About Me

Social Links