ERAIA - Enabling Intelligence Data Pipelines for IoT-based Application Systems
Aitor Hernández, Bin Xiao, Valentin Tudor · 2020
Establishing upon the connectivity layer provided by Internet of Things (IoT) platforms, modern industries are moving towards management and computation solutions which enable Artificial Intelligence (AI) services for data intensive applications. This raises two important challenges: first, the information carried by data should be refined and prepared for the various AI algorithms via data processing pipelines and; second, a distributed orchestration solution for data and AI computation resources featuring with migration capabilities is required to support the refining process. In order to address these challenges, this paper introduces ERAIA, an actor-based framework which provides a novel basis to build intelligence and data pipelines. ERAIA facilitates the deployment and migration of distributed AI computations for heterogeneous and dynamic IoT scenarios. An implementation description is accompanied by relevant performance evaluations to demonstrate the flexibility and scalability of the solution. ERAIA provides an interface to expand the scope of existing IoT systems as Application Enablement Platform (AEP), which hence accelerates the development of AI-based IoT solutions.