Education, Science, Technology, Innovation and Life
Open Access
Sign In
  1. Home
  2. Journals
  3. Automation and Machine Learning
  4. Vol 7, Issue 1, 2026

Design of an End-Cloud Collaborative Urban Security Robot System Based on Large Model

Download as PDF

DOI: 10.23977/autml.2026.070115 | Downloads: 11 | Views: 100

Author(s)

Saibo Li 1, Jiatong Lv 1, Wenxuan Zhang 1, Jianbin Feng 1, Xinning Zhang 1, Xinghai Wu 1, Maoxiang Chu 1

Affiliation(s)

1 University of Science and Technology Liaoning, Anshan, 114051, Liaoning, China

Corresponding Author

Maoxiang Chu

ABSTRACT

Aiming at the pain points of slow response, poor generalization ability of target recognition and high labor cost in traditional urban security systems in complex environments, this paper designs and implements an end-cloud collaborative security robot system driven by multimodal large model. The system adopts a three-terminal collaborative architecture of cloud decision-making, edge scheduling and end execution. The cloud deploys DINO-X Pro multimodal large model to realize accurate recognition of arbitrary semantic targets by its zero-shot learning ability. The edge side takes RDK-X5 as the main control core to build a three-level cascade framework of large detection, small detection and tracking, which solves the contradiction between large model inference delay and end-side computing power limitation. The end realizes high-precision intervention through hexapod chassis and GD32 shooting control board. The experimental data show that the system tracking frame rate is stable at 29.9 FPS, and the shooting hit rate reaches 75%, which is 55% higher than the traditional incremental PID algorithm. It has significant application value in the field of smart city dynamic security.

KEYWORDS

Smart city; Security robot; Large model; End-cloud collaboration; RBF interpolation algorithm; Zero-shot detection

CITE THIS PAPER

Saibo Li, Jiatong Lv, Wenxuan Zhang, Jianbin Feng, Xinning Zhang, Xinghai Wu, Maoxiang Chu. Design of an End-Cloud Collaborative Urban Security Robot System Based on Large Model. Automation and Machine Learning (2026). Vol. 7, No. 1, 119-125. DOI: http://dx.doi.org/10.23977/autml.2026.070115.

REFERENCES

[1] Zijian L. Integration of Artificial Intelligence and Internet of Things to Construct Urban Security Monitoring System[J].Journal of Safety Science and Engineering,2024,1(4):DOI:10.62517/JSSE.202408405.
[2] ZhengJ ,JeonS ,YangX .GRDATFusion: A gradient residual dense and attention transformer infrared and visible image fusion network for smart city security systems in cloud and fog computing[J].Expert Systems,2024,42(2):e13685.DOI:10.1111/EXSY.13685.
[3] GuoS ,WuK ,JeonS , et al.IETAFusion: An illumination enhancement and target‐aware infrared and visible image fusion network for security system of smart city[J].Expert Systems,2024,42(1):e13538.DOI:10.1111/EXSY.13538.
[4] Sun T ,Xia Y ,Gan Y .Discussion on Integration of Urban Video Surveillance System[J].Procedia Engineering,2011,15(C):3255-3259.DOI:10.1016/j.proeng.2011.08.611.
[5] Yihan K ,Xuanang R ,Renjie Z .Deep learning for image super-resolution and its application in urban security[C]//The Storm King School (United States);Chengdu Foreign Language School (China);Chengdu Tanghu Foreign Language School (China),2022:DOI:10.1117/12.2639314. 

Subscription
E-Mail Alert
Most popular papers
Call for Special Issue
Publication Ethics & OA Statement
Plagiarism Policy
Contact Us
Downloads: 5209
Visits: 261692

Sponsors, Associates, and Links

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.