research |development
software architecture for machine learning applications
feasibility analysis and design of a platform for a modular building automation system based on ai
in order to be able to use the AI in WICKIE M sensibly in the customer environment in the future, it should be structured, expandable, hardware-independent and scalable. since the current implementation comes from a feasibility analysis, the aspect of a suitable software architecture has not been considered (component implementation). as a result, not all software requirements are met.
as the project progressed, extensions and optimizations of the system were also designed. increasing functionality goes hand in hand with further, new components within the ai. If no fixed implementation scheme is defined, the ai complexity increases enormously. the aim of this work is therefore to define a suitable software architecture that best suits the system's requirements.