UC4 - Smart Traffic Management

The trials confirmed that AI-based traffic monitoring over 5G SA can reliably process video feeds from multiple high-resolution cameras, achieving downlink throughput above 500 Mbps, uplink of 63.5 Mbps, and end-to-end latency of 20 ms. Application-layer throughput ranged from 11.3 Mbps for a single camera to 67.6 Mbps for six cameras, with round-trip latency dominated by edge processing (216–616 ms). The system successfully detected traffic flow, pedestrian and vehicle behavior, and triggered alerts for hazardous situations, validating the software architecture and AI inference pipeline under realistic urban conditions. Future 6G deployments should target above 10 Gbps uplink per cell to accommodate city-wide high-resolution streams, sub-millisecond latency for instantaneous traffic adjustments, and distributed edge compute for real-time analytics. AI-driven dynamic orchestration, self-healing, granular network slicing, and embedded security will be critical to ensure scalability, low latency, reliability, and compliance with privacy regulations for broader urban deployment.{: .text-justify}