Energy-Efficient Processing Units for the IoT ERA: A New Paradigm Sustainable Edge AI for Real Time Data Optimization
Omoikhefe Aienloshan *
Department of Business Analytics, Hult International Business School, San Francisco, CA, California, United States.
*Author to whom correspondence should be addressed.
Abstract
In the current dynamic environment, the Internet of Things (IoT) stands as one of the critical fundamentals that provide ingenuity to the connection of the numerous devices that transact real-time data. From smart thermostats to industrial sensors, these are ingredefining everything right from the home to the city. However such advance increases the number of challenges especially in terms of energy consumption process analysis and system sustainability looking at the facts as there are millions of connected devices brought by IoT. Since these devices create large volumes of data, the initial cloud computing models are less suitable because of latency problems and power consumption. This has forced the emergence of edge computing where the processing of data is done nearer to the source hence minimal need to communicate a lot with the cloud. Due to the need for interconnecting different large data centres, the energy requirements of this new architecture are becoming a challenge. Edge Artificial Intelligence (AI) combining edge computing with AI is a challenging idea since it has massive benefits in terms of power use while processing power is highly mandatory for the IoT generation. This paper aims to examine the potential of the green processing elements in the IoT architecture with reference to the potential of sustainable edge learning AI enhancement on the efficiency and functionality of IoT devices. This will be done after a critical discussion of the current issues affecting AI, the newly developed solutions, and the future evolution of edge artificial intelligence for IoT sustainability.
Keywords: Artificial Intelligence (AI), energy-efficient processing, AI solutions, IoT applications