Functional Description of AI-Engines within an AI-Architecture

Mike Reichardt, Dennis Krummacker, Shreya Tayade

In: 2020 Workshop on Next Generation Networks and Applications (NGNA 2020). Workshop on Next Generation Networks and Applications (NGNA-2020) Technische Universität Kaiserslautern 2020.


In this paper, we propose a way to describe and defineArtificial Intelligence (AI) processing entities, so called AI-Engines.Such AI-Engines are arranged in an isolated manner, assignedto fulfill a well-distinguished task, while collaborating within acomprehensive AI-Architecture. Therefore, we use methods fromthe product development for defining the main functions of everyAI-Engine for clarity and highlighting the general purpose of theAI-Engine within the whole system.Furthermore, we describe the interfaces between the AI-Engines that have well-defined APIs. Along with the interfaces,we distinguish the functionalities of AI-Engine and discuss itslimitations. We also present our findings and advantages of thedescription methods and give a possible way to proceed onwards.

Weitere Links

German Research Center for Artificial Intelligence
Deutsches Forschungszentrum für Künstliche Intelligenz