Jefkine Kafunah

Jefkine Kafunah


Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

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

Representative Publications

View All Publications
7 June 2023 in IEEE Transactions on Industrial Informatics
Verified
Journal Article
Closed access
JKJefkine Kafunah
Jefkine Kafunah
MAMuhammad Intizar Ali
Muhammad Intizar Ali
John Gerard BreslinJohn Gerard Breslin
John Gerard Breslin
With the advent of Industry 4.0 (I4.0) leading to the proliferation of industrial process data, deep learning (DL) techniques have become instrumental in developing intelligent fault diagnosis (FD) applications. However, despite their potentially superior process monitoring capabilities, DL-based FD models are poorly calibrated and generate point estimate predictions without the associated uncertainty estimates. For DL-based FD models, accurate predictive uncertainty estimates from well-calibrated models are essential in ensuring industrial process safety and reliability. This article proposes ensemble-to-distribution (E2D), an uncertainty-aware combination method for quality monitoring FD based on an ensemble of deep neural networks. First, E2D addresses safety by providing accurate uncertainty estimates on model predictions, enabling informed decision-making to minimize operational risks. Second, E2D improves model performance on out-of-distribution detection tasks to facilitate deployments in the real world. Third, E2D is a post hoc application, implementable at inference time, and compatible with diverse pretrained models. Finally, to demonstrate the effectiveness of E2D, we explore the problem of monitoring the stability of industrial processes and product quality using case studies on the steel plates faults and APS failure at Scania trucks datasets.

28 November 2023 in IEEE Access
Verified
Journal Article
Open access
JKJefkine Kafunah
Jefkine Kafunah
PVPriyanka Verma
Priyanka Verma
MAMuhammad Intizar Ali
Muhammad Intizar Ali
John Gerard BreslinJohn Gerard Breslin
John Gerard Breslin
The emergence of Industry 4.0 (I4.0) has transformed modern-day factories into expansive industrial sites through rapid automation and increased access to real-time data. Deep learning (DL) approaches possessing superior capabilities for intelligent data-driven fault diagnosis (FD) have become critical in ensuring process safety and reliability in these industrial sites. However, DL-based FD applications trained exclusively on in-distribution (ID) process data face challenges in the wake of previously unseen out-of-distribution (OOD) data in the real world. This paper addresses the challenge of OOD detection for DL-based FD models by generating synthetic OOD data to simulate real-world anomalies not present in the training set. Our research aims to enhance the robustness of DL-based FD systems against OOD data. We propose Manifold Guided Sampling (MGS), a data-driven method for generating synthetic OOD samples from the ID data-supporting manifold estimated through a deep generative model. Synthetic OOD data from MGS enhances the DL-based FD model capacity for estimating the prediction uncertainty by including insights from OOD data, thereby delivering safe and reliable DL-based FD models for real-world industrial process monitoring. Additionally, the MGS algorithm maintains ID data feature space as a reference point during data generation to ensure the resulting synthetic OOD data is realistic. We analyze the effectiveness of MGS through experiments conducted on the steel plates faults dataset and demonstrate that augmenting training data with synthetic OOD data from MGS enhances DL-based FD model performance in OOD detection tasks and robustness against dataset distributional shifts. The findings underscore the effectiveness of utilizing synthetic MGS-generated OOD data in scenarios where real-world OOD data is limited, enabling better generalization and more reliable fault detection in practical applications.

20 October 2021 in Applied Sciences
Verified
Journal Article
Open access
PDF file
JKJefkine Kafunah
Jefkine Kafunah
MAMuhammad Intizar Ali
Muhammad Intizar Ali
John Gerard BreslinJohn Gerard Breslin
John Gerard Breslin
Over the recent years, Industry 4.0 (I4.0) technologies such as the Industrial Internet of Things (IIoT), Artificial Intelligence (AI), and the presence of Industrial Big Data (IBD) have helped achieve intelligent Fault Detection (FD) in manufacturing. Notably, data-driven approaches in FD apply Deep Learning (DL) techniques to help generate insights required for monitoring complex manufacturing processes. However, due to the ratio of instances where actual faults occur, FD datasets tend to be imbalanced, leading to training challenges that result in inefficient DL-based FD models. In this paper, we propose Dual Logits Weights Perturbation (DLWP) loss, a method featuring weight vectors for improved dataset generalization in FD systems. The weight vectors act as hyperparameters adjusted on a case-by-case basis to regulate focus accorded to individual minority classes during training. In particular, our proposed method is suitable for imbalanced datasets from safety-related FD tasks as it generates DL models that minimize false negatives. Subsequently, we integrate human experts into the workflow as a strategy to help safeguard the system. A subset of the results, model predictions with uncertainties exceeding a preset threshold, are considered a preliminary output subject to cross-checking by human experts. We demonstrate that DLWP achieves improved Recall, AUC, F1 scores.

15 October 2025 in Science of The Total Environment
Verified
Journal Article
Open access
NNNoel Ndlovu
Noel Ndlovu
DSDavid Styles
David Styles
JKJefkine Kafunah
Jefkine Kafunah
MNMukund Narayanan
Mukund Narayanan
LQLuis Felipe Quiroz
Luis Felipe Quiroz
CSCharles Spillane
Charles Spillane
Global demand for ruminant milk-based products is increasing, contributing to increases in associated environmental impacts. Yet, most efforts to reduce the total environmental impact of dairy production are based on livestock breeding and manipulation of feed and manure systems. Despite the critical need for low-environmental impact breeding, life cycle assessment (LCA) has not been significantly utilized for forage and livestock trait prioritization. By combining LCA and machine learning, we inventory reductions in footprints of grass-fed dairy systems achievable by selective breeding. To measure the environmental impacts of traits and other farm changes, a pasture-centric dairy system representative of Ireland was modelled using the GOBLIN model. Milk produced in ryegrass-based dairy systems in Ireland was associated with environmental impacts of 1.08 kg CO2-eq (global warming), 0.0066 kg PO4-eq (eutrophication), 0.013 kg SO2-eq (acidification), and 1.62 MJ-eq (fossil resource depletion) per kg fat- and protein-corrected milk (FPCM), translating to annual per-hectare loads of 9281.21 kg CO2-eq, 56.75 kg PO4-eq, 108.21 kg SO2-eq, and 14,025.16 MJ-eq, respectively. Using a machine learning model trained on farm-level LCA outputs, complementarities and trade-offs were revealed across 10,000 graduated scenarios analyzed. The XGBoost Regressor achieved an outstanding R2 value of 99 % in estimating the LCA impacts. Principal component analysis and explainable artificial intelligence analyses identified dry matter digestibility, crude protein, and chemical nitrogen use as key drivers of environmental impacts in dairy systems. By optimizing input parameters, the environmental impacts of grass-based milk can be substantially reduced by breeding for 'LCA-designed ideotypes'. The optimal ryegrass ideotype identified for grass-based dairy systems can reduce; (a) global warming potential by 36.7 %, (b) acidification potential by 31 %, (c) eutrophication potential by 29 %, and (d) fossil resource depletion by 11 %, compared to current levels. We conclude that the environmental performance of ryegrass-based dairy systems can be substantially increased by new LCA-designed forage ideotypes. A more systems-based approach to livestock and forage breeding is therefore needed, as a sole focus on low-footprint livestock may overlook critical gains from forage grass improvement.