A5062558456- Anomaly Detection Techniques and Applications
- Time Series Analysis and Forecasting
- Software System Performance and Reliability
- IoT and Edge/Fog Computing
- Data Visualization and Analytics
- Non-Destructive Testing Techniques
- Traffic Prediction and Management Techniques
- Network Security and Intrusion Detection
- Data Management and Algorithms
- Elevator Systems and Control
- Water Systems and Optimization
- Manufacturing Process and Optimization
- Fault Detection and Control Systems
- Intravenous Infusion Technology and Safety
- Machine Fault Diagnosis Techniques
- Digital Transformation in Industry
- Quality and Safety in Healthcare
- Semantic Web and Ontologies
- Reliability and Maintenance Optimization
- Water Quality Monitoring Technologies
- IoT and GPS-based Vehicle Safety Systems
Ghent University
2020-2024
Anomalies and faults can be detected, their causes verified, using both data-driven knowledge-driven techniques. Data-driven techniques adapt internal functioning based on the raw input data but fail to explain manifestation of any detection. Knowledge-driven inherently deliver cause that were detected require too much human effort set up. In this paper, we introduce FLAGS, Fused-AI interpretabLe Anomaly Generation System, combine in one methodology overcome limitations optimize them limited...
The wide adoption of smart machine maintenance in manufacturing is blocked by open challenges the Industrial Internet Things (IIoT) with regard to robustness, scalability and security. Solving these uttermost importance mission-critical industrial operations. Furthermore, effective application predictive requires well-trained learning algorithms which on their turn require high volumes reliable data. This paper addresses both presents Smart Maintenance Living Lab, an test research platform...
Remaining useful life is of great value in the industry and a key component Prognostics Health Management (PHM) context Predictive Maintenance (PdM) strategy. Accurate estimation remaining (RUL) helpful for optimizing maintenance schedules, obtaining insights into degradation, avoiding unexpected breakdowns. This paper presents methodology creating health index models with monotonicity semi-supervised approach. The indexes are then used enhancing models. evaluated on two bearing datasets....
In industry, dashboards are often used to monitor fleets of assets, such as trains, machines or buildings. industrial fleets, the vast amount sensors evolves continuously, new sensor data exchange protocols and formats introduced, visualization types may need be introduced existing dashboard visualizations updated in terms displayed sensors. These requirements motivate development dynamic dashboarding applications. These, opposed fixed-structure applications, allow users create at will do...
Manufacturers can plan predictive maintenance by remotely monitoring their assets. However, to extract the necessary insights from data, they often lack sufficiently large datasets that are labeled human experts. We suggest combining knowledge-driven and unsupervised data-driven approaches tackle this issue. Additionally, we present a dynamic dashboard automatically visualizes detected events using semantic reasoning, assisting experts in revision correction of event labels. Captured label...
Companies are increasingly gathering and analyzing time-series data, driven by the rising number of IoT devices. Many works in literature describe analysis systems built using either data-driven or semantic (knowledge-driven) techniques. However, little to no hybrid combinations these two. Dyversify, a collaborative project between industry academia, investigated how event anomaly detection can be performed on data such setting. We proof-of-concept platform, microservice architecture ensure...
Data visualization recommendation aims to assist the user in creating visualizations from a given dataset. The process of appropriate requires expert knowledge available data model as well dashboard application that is used. To relieve requiring this and manual numerous or dashboards, we present context-aware recommender system (VisCARS) for monitoring applications automatically recommends personalized user, based on they are task trying achieve. Through graph-based approach, about included...
Abstract Up to 30% of all drinking water is wasted due leaks in distribution networks (WDNs). In times drought and shortage, wasting so much has a considerable environmental financial cost. this paper, we present microservice architecture for leak localization WDNs, where heterogeneous sources data consisting sensor measurements, Geographic Information System (GIS), Customer Relationship Management (CRM) are used feed monitoring solution which combines hybrid data-driven model-based...