Skip to main navigation Skip to search Skip to main content

AI-Powered IoT System at the Edge

  • Yiran Chen
  • , Ang Li
  • , Huanrui Yang
  • , Tunhou Zhang
  • , Yuewei Yang
  • , Hai Li
  • , Suman Banerjee
  • , Miroslav Pajic

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The proliferation of low-cost and low-power IoT devices are constantly generating gigabytes data at the network edge. Bridging AI with IoT is a natural option to unleash the data on devices. AI-powered IoT systems can boost many novel applications and services with impressive performance. In this paper, we identify four key challenges for developing AI-powered IoT systems, including interpretability, efficiency, heterogeneity, and trustworthiness. We also summarize our proposed solutions to tackle each of these challenges. Finally, we discusses the open challenges and future research directions on designing inter-pretable, efficient, heterogeneous, and trustworthy IoT systems.

Original languageEnglish (US)
Title of host publicationProceedings - 2021 IEEE 3rd International Conference on Cognitive Machine Intelligence, CogMI 2021
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages242-251
Number of pages10
ISBN (Electronic)9781665416214
DOIs
StatePublished - 2021
Event3rd IEEE International Conference on Cognitive Machine Intelligence, CogMI 2021 - Virtual, Online, United States
Duration: Dec 13 2021Dec 15 2021

Publication series

NameProceedings - 2021 IEEE 3rd International Conference on Cognitive Machine Intelligence, CogMI 2021

Conference

Conference3rd IEEE International Conference on Cognitive Machine Intelligence, CogMI 2021
Country/TerritoryUnited States
CityVirtual, Online
Period12/13/2112/15/21

Keywords

  • Artificial Intelligence
  • Edge Computing
  • IoT
  • Machine Learning

ASJC Scopus subject areas

  • Computer Vision and Pattern Recognition
  • Cognitive Neuroscience
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'AI-Powered IoT System at the Edge'. Together they form a unique fingerprint.

Cite this