Towards AI powered manufacturing services, processes, and products in an edge-to-cloud-knowlEdge continuum for humans [in-the-loop]
Artificial intelligence (AI) is the software engine for the fourth industrial revolution that is changing the way we live and work. However, the complex technologies and the lack of skilled talent are barriers to progressing AI and thus increasing product quality and business sustainability. The EU-funded knowlEdge project will address the need for new AI solutions that are agile, reusable, distributed, scalable, accountable, secure, standardised and collaborative. The proposed new framework will ensure the secure management of distributed data and facilitate knowledge exchange. To achieve its goal, the project will combine innovative technologies from data management, data analytics and knowledge management.
Gaussian processes for automatic and interpretable anomaly detection
This research project aims to explore Gaussian processes for efficient detection and interpretation of anomalies in multivariate time series data. In particular, unsupervised Gaussian processes will be investigated and further developed in order to identify, understand and resolve underlying correlations and anomalies. In order to learn Gaussian process models in a scalable and real-time manner, we intend to develop new streaming algorithms, which will be implemented in an open source manner and with reference to industrial standards, and tested in application-oriented scenarios, together with industry partners.
Efficient Ptolemaic Indexing
Concomitant with the rapid growth of heterogeneous data, the demand for efficient and scalable data access increases. Ptolemaic Access Methods provide a domain-agnostic approach for indexing and accessing complex data spaces based on metric similarity models. While initial studies have already demonstrated the efficiency of this comparatively young indexing method in various data-intensive domains, the fundamentals of this approach are largely unexplored. Questions concerning the approximation of distances in metric and ptolemaic data spaces, the geometry of ptolemaic queries, as well as the interaction of different lower bounding methods are currently considered to be not sufficiently answered. The aim of this research project is to investigate the fundamentals of ptolemaic and metric access methods and to methodically advance the findings obtained in order to demonstrate the performance of this class of access methods for indexing large, complex data spaces. This research project thus pursues the overall goal of advancing the development of efficient data technologies for exploring digital data resources.