Duncan Kibet

Mr. Duncan Kibet

orcid

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
17 December 2024 in Sensors
Verified
Journal Article
Open access
PDF file
DKDuncan Kibet
Duncan Kibet
MSMin-Seop So
Min-Seop So
HKHahyeon Kang
Hahyeon Kang
YHYongSu Han
YongSu Han
Jong-Ho ShinJong-Ho Shin
Jong-Ho Shin
In human activity recognition, accurate and timely fall detection is essential in healthcare, particularly for monitoring the elderly, where quick responses can prevent severe consequences. This study presents a new fall detection model built on a transformer architecture, which focuses on the movement speeds of key body points tracked using the MediaPipe library. By continuously monitoring these key points in video data, the model calculates real-time speed changes that signal potential falls. The transformer’s attention mechanism enables it to catch even slight shifts in movement, achieving an accuracy of 97.6% while significantly reducing false alarms compared to traditional methods. This approach has practical applications in settings like elderly care facilities and home monitoring systems, where reliable fall detection can support faster intervention. By homing in on the dynamics of movement, this model improves both accuracy and reliability, making it suitable for various real-world situations. Overall, it offers a promising solution for enhancing safety and care for vulnerable populations in diverse environments.

23 November 2023 in Machines
Verified
Journal Article
Open access
PDF file
MSMin-Seop So
Min-Seop So
DKDuncan Kibet
Duncan Kibet
TWTae Kyeong Woo
Tae Kyeong Woo
SKSeong-Joon Kim
Seong-Joon Kim
Jong-Ho ShinJong-Ho Shin
Jong-Ho Shin
Coal has been used as the most commonly energy source for power plants since it is relatively cheap and readily available. Thanks to these benefits, many countries operate coal-fired power plants. However, the combustion of coal in the coal-fired power plant emits pollutants such as sulfur oxides (SOx) and nitrogen oxides (NOx) which are suspected to cause damage to the environment and also be harmful to humans. For this reason, most countries have been strengthening regulations on coal-consuming industries. Therefore, the coal-fired power plant should also follow these regulations. This study focuses on the prediction of harmful emissions when the coal is mixed with high-quality and low-quality coals during combustion in the coal-fired power plant. The emission of SOx and NOx is affected by the mixture ratio between high-quality and low-quality coals so it is very important to decide on the mixture ratio of coals. To decide the coal mixture, it is a prerequisite to predict the amount of SOx and NOx emission during combustion. To do this, this paper develops a deep neural network (DNN) model which can predict SOx and NOx emissions associated with coal properties when coals are mixed. The field data from a coal-fired power plant is used to train the model and it gives mean absolute percentage error (MAPE) of 7.1% and 5.68% for SOx and NOx prediction, respectively.

4 August 2023 in Processes
Verified
Journal Article
Open access
PDF file
DKDuncan Kibet
Duncan Kibet
Jong-Ho ShinJong-Ho Shin
Jong-Ho Shin
In this research work, an approach using You Only Look Once version three (YOLOv3)-TensorFlow for abalone detection and Deep Simple Online Real-time Tracking (DeepSORT) for abalone tracking in conveyor belt systems is proposed. The conveyor belt system works in coordination with the cameras used to detect abalones. Considering the computational effectiveness and improved detection algorithms, this proposal is promising compared to the previously proposed methods. Some of these methods have low effectiveness and accuracy, and they provide an incorrect counting rate because some of the abalones tend to entangle, resulting in counting two or more abalones as one. Conducting detection and tracking research is crucial to achieve modern solutions for small- and large-scale fishing industries that enable them to accomplish higher automation, non-invasiveness, and low cost. This study is based on the development and improvement of counting analysis tools for automation in the fishing industry. This enhances agility and generates more income without the cost created by inaccuracy.

7 December 2025 in Energies
Verified
Journal Article
Open access
PDF file
DKDuncan Kibet
Duncan Kibet
MSMin-Seop So
Min-Seop So
Jong-Ho ShinJong-Ho Shin
Jong-Ho Shin
Solar power forecasting is important for energy management and grid stability, yet many deep learning studies use a large set of meteorological and time-based variables because of the belief that more inputs improve model performance. In practice, a large feature set can introduce redundancy, increase computational effort, and reduce clarity in model interpretation. This study examines whether dependable forecasting can be achieved using only the most influential variables, presenting a minimal feature deep learning approach for short term prediction of solar power. The objective is to evaluate a Transformer model that uses only two key variables, solar irradiance and soil temperature at a depth of ten centimetres. These variables were identified through feature importance analysis. A real world solar power dataset was used for model development, and performance was compared with RNN, GRU, LSTM, and Transformer models that use the full set of meteorological inputs. The minimal feature Transformer reached a Mean Absolute Error of 1.1325, which is very close to the result of the multivariate Transformer that uses all available inputs. This outcome shows that essential temporal patterns in solar power generation can be captured using only the strongest predictors, supporting the usefulness of reducing the size of the input space. The findings indicate that selective feature reduction can maintain strong predictive performance while lowering complexity, improving clarity, and reducing data requirements. Future work may explore the adaptability of this minimal feature strategy across different regions and environmental conditions.

30 December 2024 in Journal of Society of Korea Industrial and Systems Engineering
Verified
Journal Article
Closed access
HKHaHyeon Kang
HaHyeon Kang
MSMin Seop So
Min Seop So
DKDuncan Kibet
Duncan Kibet
Jong-Ho ShinJong-Ho Shin
Jong-Ho Shin

9 February 2026 in Sensors
Verified
Journal Article
Open access
PDF file
MSMin-Seop So
Min-Seop So
DKDuck Bong Kim
Duck Bong Kim
DKDuncan Kibet
Duncan Kibet
Jong-Ho ShinJong-Ho Shin
Jong-Ho Shin
Conventional subtractive manufacturing methods, such as cutting, often result in material waste and limitations in geometric complexity. To address these challenges, Wire Arc Additive Manufacturing (WAAM), in which components are built through successive weld bead deposition, has attracted increasing attention across various industrial fields. However, WAAM-fabricated components typically exhibit significant surface irregularities, necessitating additional post-processing that reduces overall productivity. Improving productivity therefore requires effective control and optimization of deposition parameters. This task is particularly challenging in multilayer WAAM processes, as the geometry of previously deposited layers varies with operating conditions. To address this challenge, this study proposes an AI-based framework for controlling surface roughness by rapidly identifying near-optimal process parameters in response to evolving bead geometry. A large-scale simulation dataset was generated by applying a pre-trained deep neural network (DNN) surface roughness predictor to one million bead geometry variations under 72 process parameter combinations. The resulting optimal parameter labels were used to train a classification model that recommends process conditions based on the current bead geometry. Model performance was evaluated using predictor-estimated surface roughness values, achieving Weighted Precision, Recall, and F1-score of 0.98, with an average AUC of 0.977. Five previously generated WAAM specimens were used for comparative analysis between AI-recommended and conventional process conditions using the previously developed and validated surface roughness prediction model, rather than direct physical measurements. This predictor-based feasibility analysis showed that AI-recommended conditions consistently reduced the predicted surface roughness, indicating the potential of AI-driven process optimization to improve surface quality in WAAM and reduce reliance on post-processing.