Browsing by Author "Jerye, Felix"

Now showing 1 - 2 of 2

Open Access
Exploring the challenges of enterprise integration in the cloud
(2024) Jerye, Felix
The integration of applications and services in a distributed system has long existed in enterprises and organizations. With the rise of the cloud and the utilization of various Managed Service Providers (MSPs) and Cloud Service Providers (CSPs) in addition to the own system landscape, the whole situation becomes even more complex. From the initial integration, to changes and additions, legal and business constraints, performance and cost monitoring, security and organizational structures, there are many challenges that enterprises are facing. It is therefore necessary to understand how such integrations should be realized and how they can be observed and evaluated. The aim of this thesis is to find out which integration technologies are used in the cloud, what problems they cause and how these can be overcome. Which tools are used and how do they differ? What about integration middleware in the cloud? How much added value do off-the-shelf integration components offer? How can integrations be observed and properly evaluated? In order to answer these questions, a literature review and further research were carried out as well as expert interviews and a case study. Using a practical example, a data source integration into the BigData4Biz (BD4B) platform, various approaches were compared and analyzed for observability. It has been shown that a variety of integration approaches are used in real-world enterprise multicloud environments. From point-to-point integrations in microservices, to event-driven integrations with Apache Kafka, the use of Integration Platform-as-a-service (iPaaS) solutions, the deployment of more lightweight Enterprise Service Bus (ESB) successors and others can exist in the same integration flow in an enterprise system. This is often due to merger of enterprises, organizational structures, such as different teams, sometimes across countries and departments, business complexities, historical backgrounds, legal requirements or simply bad design. Something that has to be accepted as a fact to happen and worked with. Also it has been shown that it is almost impossible to observe and evaluate what happens in such a system-wide multicloud integration. Through the System Integration Watchdog, in addition to existing observability tools, it was shown that it would be possible to obtain a system-wide view of the integration by extracting the integration flows, in combination with other collected data through more standardized methods, like OpenTelemetry. It is a proof of concept to demonstrate feasibility. Further research in the area of integration in complex cloud environments should focus on observation in order to obtain a better data basis for statements about the integrations. The integration tools, which are significantly more heterogeneous, should also be subjected to a comprehensive examination based on features, as there is a lack of classifications and methods with the existing ones being too vague.
Open Access
What are the current capabilities for a speech interface using existing speech libraries on a mobile robot with limited hardware resources?
(2022) Jerye, Felix
Interaction with robots often requires a speech interface. Particularly robots with many concurrent processes, need a resource efficient solution. The available frameworks are often tightly coupled, making them less suitable for an deep integration. This work presents a modular prototype speech interface based on the Robot Operating System. The individual modules are interchangeable and logically separated. Special attention is paid to the weak hardware of a Raspberry Pi 4. Benchmarks are used to show that this platform is suitable for a real-time speech interface. This does not apply to all software components. Especially GPU-optimized components reach their limits and miss real-time.