Adaptive Orchestration Mechanisms for Efficient Serverless Collection of Heterogeneous Environmental Data
The rapid expansion of cyber-physical systems (CPS) has intensified the need for scalable and adaptive mechanisms to collect heterogeneous environmental data from numerous unstable external sources. Traditional serverless orchestration frameworks, while elastic and cost- efficient, lack runtime adaptability and feedback-awareness, leading to inefficiencies under dynamic API conditions. This paper presents a novel adaptive orchestration model for serverless data collection pipelines, driven by metadata configuration and continuous feedback control.