seminar • June 21, 2024
Contextual anomaly detection for complex industrial assets
Unveiling the insights of CONSCIOUS
Join us for an exclusive seminar uncovering how the CONSCIOUS project redefines anomaly detection in industrial settings, empowering you to differentiate true irregularities from contextual nuances with AI precision.
Decoding anomalies in industrial realms
CONSCIOUS is among the select AI ICON projects funded by VLAIO (The Flemish Agency for Innovation & Entrepreneurship) aimed at bridging the gap between research results in AI and their implementation in the Flemish business community.
CONSCIOUS focuses on detecting abnormal behavior in industrial machines and processes by analyzing data from sensors. Its innovative approach emphasizes contextual understanding to differentiate between explainable deviations and true anomalies. Put differently, the project teaches AI systems to identify causes of irregularities within contextually enriched data, ensuring more precise detection of genuine issues.
Context detection
Addressing the challenge of automatically discerning the contexts in which monitored assets operate.
Anomaly detection within contexts
Leveraging learned contextual information to enhance anomaly detection algorithms.
Human-in-the-loop systems
Exploring methods to enrich anomaly detection by integrating human expertise.
Concluding seminar: unveiling CONSCIOUS insights
This seminar will feature an in-depth discussion of the insights and tangible impact generated by the CONSCIOUS project through three compelling use cases spanning different industrial scenarios.
Date: June 21, 2024
Time: 12:00 PM — 05:30 PM
Location: Skyline Park Izegem
Price: Free, but registration is mandatory.
Please note that seats are limited, so make sure to register before June 14th to secure your spot.
AGENDA
Welcome with sandwiches
Introduction on CONSCIOUS: project goals and use cases
Use case by Duracell: manufacturing processes
Coffee break
Use case by Skyline Communications: networked asset management
Coffee break
Use case by Icare: machinery
Reception and networking
GET TO KNOW THE PARTNERS
Duracell is the world leader in alkaline, specialty, and rechargeable batteries of superior performance. Since the company was founded in the 1940s, it has become an iconic brand when it comes to powering your personal items with compact, long-lasting batteries. For Europe and Africa, these batteries are being produced in the heart of Europe in the manufacturing facilities in Aarschot, Flanders.
Skyline’s flagship product, DataMiner, is vital for global telecom operators, helping them manage vast networks efficiently. The complexity of these networks, with numerous interconnected devices, requires advanced tools for root cause analysis and fault detection. In response, Skyline has integrated three cutting-edge relation-learning features and an anomaly feedback mechanism into DataMiner as part of the CONSCIOUS project. The project has yielded several other valuable research outcomes ready for implementation.
The activities of the Declarative Languages and Artificial Intelligence section (DTAI, Dept CS, KU Leuven; Leuven.AI) are centered around research and education in AI and machine learning. DTAI is internationally renowned for its expertise in integrating different forms of learning and reasoning, neuro-symbolic models, probabilistic programming, time series, learning from structured data, knowledge representation and graphs, and constraint programming. DTAI applies these insights to use cases in CONSCIOUS to achieve motif discovery, segmentation and anomaly detection.
The EluciDATA Lab, part of Sirris, specializes in data processing and AI to drive innovation in the Belgian tech industry. Through the CONSCIOUS project, the Lab is tackling context-aware anomaly detection in industrial settings. These environments pose challenges due to their dynamic nature, making anomaly detection difficult. The project aims to develop precise, robust, and interpretable anomaly detection solutions by considering contextual factors. Real-world validation will be conducted across various industrial domains.
Fingerhut, F.; Verbeke, M.; Tsiporkova, E.,Unsupervised context-sensitive anomaly detection on streaming data relying on multi-view profiling, to appear in Conference proceedings of IEEE EAIS 2024, Madrid
Eghbalian A.;, Vroman A.; Klein S.; Tsiporkova, E.; Spatiotemporal context detection in distributed systems events with an application on network communication alarms, submitted to 28th International Conference on Knowledge-Based and Intelligent Information & Engineering Systems (KES 2024)
Adrem Data Lab is embedded in the department of Computer Science at the University of Antwerp. The lab builds on a long standing and internationally strong reputation ranging from the theoretical foundations to the practical application of data sciences. The lab covers many important facets required to manage, query, interpret and analyze data, in all its forms. This results in a broad and recognised excellence in areas ranging from data cleaning, data mining, machine learning, and artificial intelligence. Adrem Data Lab applies these insights to use cases in CONSCIOUS for pattern and motif discovery, context detection and anomaly detection.
I-care is a global leader in predictive and prescriptive maintenance. The AI-powered and data driven solutions predict industrial failures before anyone else. With innovative methods, products and services, I-care takes care of thousands of assets and processes of clients around the world and helps them reduce risks for safer, more productive, and more sustainable industries. Founded in 2004 and based in Belgium, I-care employs over 500 people and has a presence in 26 countries (Europe, USA, and Asia-Pacific). It was named Business of the Year 2020.
Yazzoom provides data analysis solutions to businesses, research institutions, and government bodies. Established in 2011, the Belgian company has built a solid reputation for optimizing industrial processes using their expertise in software development, computer modelling, machine learning, advanced process control, signal processing, and data mining, enhancing customer processes. Alongside consultancy, Yazzoom offers its own software products Yasense and Yanomaly for anomaly detection on industrial assets. As part of the CONSCIOUS project, Yazzoom has developed an advanced AI system capable of identifying contextual information within complex industrial data and leveraging it for anomaly detection. This system will be integrated into the Yanomaly product, elevating its anomaly detection capabilities to a new level.
Our group is a global reference in low-carbon energy and services. Together with our 97,000 employees, our customers, partners and stakeholders, we are committed to accelerate the transition towards a carbon-neutral world, through reduced energy consumption and more environmentally-friendly solutions. Inspired by our purpose (“raison d’être”), we reconcile economic performance with a positive impact on people and the planet, building on our key businesses (gas, renewable energy, services) to offer competitive solutions to our customers.
