A new Edge Architecture for AI-IoT services deployment

Artificial intelligence (AI) and Internet of things (IoT) have progressively emerged in all domains of our daily lives. Nowadays, both domains are combined in what is called artificial intelligence of thing (AIoT). With the increase of the amount of data produced by the myriad of connected things and the large wide of data needed to train Artificial Intelligence models, data processing and storage became a real challenge. Indeed, the amount of data to process, to transfer by network and to treat in the cloud have call into question classical data storage and processing architectures. Also, the large amount of data generated at the edge has increased the speed of data transportation that is becoming the bottleneck for the cloud-based computing paradigms. The post-cloud approaches using Edge computing allow to improve latency and jitter. These infrastructures manage mechanisms of containerization and orchestration in order to provide automatic and fast deployment and migration of services such as reasoning mechanisms, ontology, and specifically adapted artificial intelligence algorithms. In this paper we propose a new architecture used to deploy at edge level micro services and adapted artificial intelligence algorithms and models.