Embodied AI Through Cloud-Fog Computing: A Framework for Everywhere Intelligence

被引：0

作者：

Hu, Dongxiao ^{[1
]}

Lan, Dapeng ^{[2
]}

Liu, Yu ^{[3
]}

Ning, Jiahong ^{[4
]}

Wang, Jia ^{[1
]}

Yang, Yun ^{[5
]}

Pang, Zhibo ^{[6
]}

机构：

[1] Xian Jiaotong Liverpool Univ, Sch Adv Technol, Xian, Peoples R China

[2] Chinese Acad Sci Shenyang, Shenyang Inst Automat, Shenyang, Peoples R China

[3] Linkoping Univ, Dept Sci & Technol, Linkoping, Sweden

[4] Dalian Marine Univ, Dalian, Peoples R China

[5] Yunnan Univ, Natl Pilot Sch Software, Kunming, Yunnan, Peoples R China

[6] KTH Royal Inst Technol, Dept Intelligent Syst, Stockholm, Sweden

来源：

2024 33RD INTERNATIONAL SYMPOSIUM ON INDUSTRIAL ELECTRONICS, ISIE 2024 | 2024年

关键词：

Embodied AI; agent; cloud-fog automation; pretrained foundation models;

D O I：

10.1109/ISIE54533.2024.10595837

中图分类号：

TP39 [计算机的应用];

学科分类号：

081203 ; 0835 ;

摘要：

Embodied AI represents a crucial step towards achieving Artificial General Intelligence (AGI). The next paradigm of Embodied AI involves physical embodiment, enhanced perception capabilities, and adaptive automation. This advances the field significantly, paving the way for broader expansion. Despite the significant progress, existing computing frameworks, like local computation or cloud computing, struggle to meet the substantial demands of Embodied AI. The Cloud-Fog Embodied framework, namely based on CFA (cloud-fog automation) offers a promising solution to address these challenges. Our goal is to drive integration across multiple domains, including AI, robotics and industrial production, to tackle multifaceted challenges and seize opportunities to achieve AGI in the future.

引用

页数：4