Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Real-World Deployment of Cloud Autonomous Mobility System Using 5G Networks for Outdoor and Indoor Environments

Yufeng Yang, Minghao Ning, Keqi Shu, Aladdin Saleh, Ehsan Hashemi, Amir Khajepour

TL;DR

This paper demonstrates a real-world Cloud Autonomous Mobility (CAM) system that connects edge infrastructure sensor nodes (LiDAR and cameras) to a central cloud for global perception fusion and planning via 5G with URLLC capabilities. By deploying in both a dense urban roundabout and a hospital-like indoor corridor, the authors show improved fleet management, hazard warnings, asset tracking, and socially aware autonomous navigation. The work highlights CAM's low-latency, scalable architecture, and provides practical insights into deployment challenges and scalability for outdoor and indoor intelligent transportation systems. The results suggest that cloud-level perception fused with edge sensing can enhance safety, efficiency, and asset management in diverse mobility environments.

Abstract

The growing complexity of both outdoor and indoor mobility systems demands scalable, cost-effective, and reliable perception and communication frameworks. This work presents the real-world deployment and evaluation of a Cloud Autonomous Mobility (CAM) system that leverages distributed sensor nodes connected via 5G networks, which integrates LiDAR- and camera-based perception at infrastructure units, cloud computing for global information fusion, and Ultra-Reliable Low Latency Communications (URLLC) to enable real-time situational awareness and autonomous operation. The CAM system is deployed in two distinct environments: a dense urban roundabout and a narrow indoor hospital corridor. Field experiments show improved traffic monitoring, hazard detection, and asset management capabilities. The paper also discusses practical deployment challenges and shares key insights for scaling CAM systems. The results highlight the potential of cloud-based infrastructure perception to advance both outdoor and indoor intelligent transportation systems.

Real-World Deployment of Cloud Autonomous Mobility System Using 5G Networks for Outdoor and Indoor Environments

TL;DR

This paper demonstrates a real-world Cloud Autonomous Mobility (CAM) system that connects edge infrastructure sensor nodes (LiDAR and cameras) to a central cloud for global perception fusion and planning via 5G with URLLC capabilities. By deploying in both a dense urban roundabout and a hospital-like indoor corridor, the authors show improved fleet management, hazard warnings, asset tracking, and socially aware autonomous navigation. The work highlights CAM's low-latency, scalable architecture, and provides practical insights into deployment challenges and scalability for outdoor and indoor intelligent transportation systems. The results suggest that cloud-level perception fused with edge sensing can enhance safety, efficiency, and asset management in diverse mobility environments.

Abstract

The growing complexity of both outdoor and indoor mobility systems demands scalable, cost-effective, and reliable perception and communication frameworks. This work presents the real-world deployment and evaluation of a Cloud Autonomous Mobility (CAM) system that leverages distributed sensor nodes connected via 5G networks, which integrates LiDAR- and camera-based perception at infrastructure units, cloud computing for global information fusion, and Ultra-Reliable Low Latency Communications (URLLC) to enable real-time situational awareness and autonomous operation. The CAM system is deployed in two distinct environments: a dense urban roundabout and a narrow indoor hospital corridor. Field experiments show improved traffic monitoring, hazard detection, and asset management capabilities. The paper also discusses practical deployment challenges and shares key insights for scaling CAM systems. The results highlight the potential of cloud-based infrastructure perception to advance both outdoor and indoor intelligent transportation systems.

Paper Structure

This paper contains 13 sections, 7 figures.

Table of Contents

  1. Introduction
  2. CAM Framework Overview
  3. System Architecture
  4. Infrastructure Sensor Nodes
  5. Cloud Computing and Information Fusion
  6. Advantages of the CAM System
  7. Outdoor application
  8. Connected autonomous vehicle fleet management
  9. Road hazard and critical scenario warning
  10. Indoor application
  11. Asset Management Results
  12. Autonomous Driving Results
  13. Conclusion

Figures (7)

  • Figure 1: Cloud Autonomous Mobility System Overview
  • Figure 2: Information flow of the Cloud Autonomous Mobility System
  • Figure 3: Outdoor sensor node layout
  • Figure 4: Perception result from outdoor sensor nodes
  • Figure 5: Sensor node locations for the indoor application. Sensor Nodes 1–3 are deployed for asset management testing, while Sensor Node 4 is used for evaluating socially aware autonomous navigation
  • ...and 2 more figures