Abstract
Edge Computing (EC) is an enabling paradigm for developing technologies like the Internet of Things (IoT), 5G, online gaming, augmented reality, vehicle-to-vehicle communications, smart grids, and real-time video analytics, among others. In these scenarios, a remarkable trend is the increasing adoption of Artificial Intelligence (AI) and Machine Learning (ML) techniques to execute tasks like data classification, spam filtering, anomaly detection in network traffic for cybersecurity, real-time image classification and segmentation, driving support applications, autonomous driving vehicles, and many others. Many of these applications demand the online processing of continuous data streams, which demand (almost) real-time processing. The EC paradigm raises the opportunity to move the execution of many AI and ML operations to the network's edge, closer to the users, applications, and data sources. Many studies estimate that the energy demand to power information and communication technologies ICTs) surpassed 14% of global electricity usage in 2018. Moreover, growing around 6-8% per year, it will rise above 20% by 2030 in a non-pessimistic scenario. In this project, we will develop novel strategies for efficient, scalable and accurate execution of AI and ML operations on edge computing resources. The objective is to optimize AI and ML operations for the analysis of data streams by reducing processing latency, energy consumption, and financial costs in edge computing environments. (AU)
| Articles published in Agência FAPESP Newsletter about the research grant: |
| More itemsLess items |
| TITULO |
| Articles published in other media outlets ( ): |
| More itemsLess items |
| VEICULO: TITULO (DATA) |
| VEICULO: TITULO (DATA) |