Rakuten and T-Mobile Show Practical AI-RAN Progress
The article highlights Rakuten Mobile and T-Mobile as examples of AI-RAN moving from hype to implementation. Rakuten is using its virtualized open RAN base for automated operations, reporting more than 20% energy savings and working with Intel on edge inference alongside vRAN workloads. T-Mobile is adding AI to its self-organizing network, with tools that adjust neighboring cells after outages and forecast capacity for major events.
Rakuten Mobile’s work builds on its virtualized open RAN base, using closed-loop automation that TM Forum has validated at level four. The operator reports over 20% energy savings and is collaborating with Intel to run AI inference next to vRAN workloads, avoiding a full backhaul to centralized data centers.
T-Mobile’s US approach adds intelligence to its self-organizing network. AutoPilot uses intent-based automation to reconfigure neighboring cells when sites go down, while Dynamic CX forecasts demand and adjusts capacity for large events. T-Mobile says AutoPilot halves adjustment time; its SON made 30,000 antenna changes during Winter Storm Fern.
These developments may affect mobile users through more resilient coverage and faster recovery after outages, and could lower operators’ energy costs, potentially influencing service pricing or sustainability. Network engineers and edge-computing vendors may see changed roles as automation handles more routine optimization. However, benefits depend on deployment scale and guardrails, so impacts could vary by region and operator.