NTT DOCOMO has become the first mobile operator in Japan to deploy Nokia’s AI-powered MantaRay AutoPilot platform for automated mobile network optimization. The system entered commercial operation on June 19 and also marks what DOCOMO says is the world’s first deployment of MantaRay AutoPilot on a public cloud to optimize a live commercial mobile network. The deployment builds on DOCOMO’s earlier implementation of Nokia’s MantaRay SON in late 2025, extending automation beyond self-organizing network functions to AI-driven autonomous optimization.
Unlike traditional self-organizing network (SON) platforms that require engineers to manually design optimization parameters and policies, MantaRay AutoPilot allows operators to define high-level network “intents,” such as minimum throughput targets or quality objectives. The AI continuously analyzes network performance, traffic loads, and base station conditions before automatically generating optimization policies and execution schedules. Working alongside MantaRay SON, the platform can complete an optimization cycle in as little as 15 minutes, compared with previous daily optimization cycles. DOCOMO said the continuous AI-driven process enables more responsive optimization based on changing congestion patterns across different locations and times of day, helping deliver more consistent mobile broadband performance.
DOCOMO deployed the platform using public cloud infrastructure, allowing rapid rollout without waiting for dedicated hardware procurement. The operator also plans to integrate MantaRay AutoPilot with additional cloud-based AI platforms over time. DOCOMO said the deployment supports its objective of reaching TM Forum Autonomous Networks Level 4, where AI predicts network conditions and autonomously manages operations with minimal human intervention. The company will continue evaluating the platform in commercial operation as it expands AI-driven network management across its nationwide mobile network.
• First commercial deployment of Nokia MantaRay AutoPilot in Japan.
• First reported deployment of MantaRay AutoPilot optimizing a commercial mobile network from a public cloud.
• AI replaces manual parameter design with intent-based optimization.
• Optimization cycles reduced from approximately one day to as little as 15 minutes.
• AI continuously analyzes network conditions and automatically updates optimization policies.
• Platform operates alongside Nokia MantaRay SON in DOCOMO’s multi-vendor 5G network.
• Public cloud deployment accelerates rollout without dedicated hardware lead times.
• Supports DOCOMO’s roadmap toward TM Forum Autonomous Networks Level 4.
• Future plans include integration with additional cloud AI platforms.
“Our deployment of MantaRay AutoPilot enables AI to continuously optimize our mobile network based on service intents while significantly reducing manual operational work. This represents another step toward highly autonomous network operations,” DOCOMO said.
🌐 Analysis
This deployment illustrates how telecom AI is evolving beyond analytics and operational dashboards into closed-loop network control. Rather than simply recommending configuration changes, AI systems are increasingly making optimization decisions directly based on operator-defined intents, reducing engineering effort while allowing networks to respond much more quickly to changing traffic patterns.
The announcement also reflects two broader industry trends: cloud-native network operations and the industry’s push toward TM Forum Autonomous Networks Level 4. Nokia, Huawei, Ericsson and other major infrastructure vendors are all investing heavily in intent-based networking, AI agents and autonomous operations platforms as operators seek to improve network efficiency while managing increasingly complex 5G and future 6G infrastructures.
