Posts by Collection

portfolio

Portfolio item number 1

Short description of portfolio item number 1

Portfolio item number 2

Short description of portfolio item number 2

publications

Quality Assessment of Welding using Regression Analysis of Biomechanical Data

Published in 2024 Moratuwa Engineering Research Conference (MERCon) (IEEE), 2024

The widespread integration of wearable devices in various fields has paved the way for numerous novel applications in medicine, rehabilitation, sports, and industry, etc. through the data acquisition, and analysis of hand gestures and gait patterns. This study aims at a comprehensive exploration of a psychomotor skill, i.e., welding using a wearable device to capture hand movements. After the data collection, this study is focused on assessing the welding quality parameters such as symmetry and equal distribution of a welding path. For this, several machine learning and deep learning algorithms were trained in order to obtain the above welding quality parameters using biomechanical data which was captured using four IMU sensors placed on the hand of the worker when performing the welding task. From our results, it is evident that it is possible to use regression analysis to predict the welding quality parameters including equal distribution of a welding path with minimum error.

Download Paper

Comparison of Appliance Signature Classification Methods for Non-Intrusive Load Monitoring

Published in 2024 Moratuwa International Conference on Electrical Engineering (EECon) (IEEE), 2024

Non-Intrusive Load Monitoring (NILM) is a vital technique for disaggregating individual appliance loads from aggregated household energy consumption data. In this study, we utilize the PLAID dataset to conduct a comprehensive review of various classification methods employed in NILM systems. We focus on traditional machine learning approaches such as support vector machines (SVM), k-nearest neighbors (KNN), and random forests (RF), as well as modern deep learning models like multi-layer perceptrons (MLP) and convolutional neural networks (CNN). Our analysis examines the effect of different sampling frequencies on the performance of these models, highlighting the strengths and limitations of each under varying conditions. The insights provided in this paper offer valuable guidance for researchers and practitioners in selecting the most effective classification techniques for their specific NILM applications.

Download Paper

Estimation of Angle of Significance in Human Joint Angles Using Extended Kalman Filter and Principal Component Analysis (Unpublished)

Published in 2025 IEEE Applied Sensing Conference (APSCON) (unpublished), 2025

Human Motion Analysis (HMA) is a pivotal multidisciplinary field that leverages advanced signal processing and machine learning techniques to interpret human movement patterns. This paper explores a newly developed wireless wearable device that consists of Inertial Measurement Unit (IMU) sensors along with its usage in the estimation of human joint angles using a combination of Extended Kalman Filter (EKF) and Principal Component Analysis (PCA). The integration of EKF, a dynamic algorithm adept at estimating joint angles while accounting for noise, and PCA, a technique for dimensionality reduction and principal component identification, enhances the precision of motion analysis. Additionally, the results obtained in this study are validated using YOLOv7, ensuring the reliability and accuracy of the joint angle estimations. Hence, the study explains the collective effort of the design of the IMU-based wearable device, data collection, analysis, and validation of the joint angle estimations highlighting the accuracy and applicability of results in various domains, including medical diagnostics, sports performance, and industrial applications.

Performance Benchmarking of Psychomotor Skills Using Wearable Devices: An Application in Sport

Published in IEEE Access, 2025

Mastering psychomotor skills, such as those essential in sports, rehabilitation, and professional training, often requires a precise understanding of motion patterns and performance metrics. This study proposes a versatile framework for optimizing psychomotor learning through human motion analysis. Utilizing a wearable IMU sensor system, the motion trajectories of a given psychomotor task are acquired and then linked to points in a performance space using a predefined set of quality metrics specific to the psychomotor skill. This enables the identification of a benchmark cluster in the performance space, which represents a group of reference points that define optimal performance across multiple criteria, allowing correspondences to be established between the performance clusters and sets of trajectories in the motion space. As a result, common or specific deviations in the performance space can be identified, enabling remedial actions in the motion space to optimize performance. A thorough validation of the proposed framework is done in this paper using a Table Tennis forehand stroke as a case study. The resulting quantitative and visual representation of performance empowers individuals to optimize their skills and achieve peak performance.

Download Paper