Press Releases

June 22, 2026

DOCOMO Becomes First in Japan to Deploy Nokia's AI-powered MantaRay AutoPilot for Automated Network Optimization

  • Print

TOKYO, JAPAN, June 22, 2026 --- NTT DOCOMO, INC. announced today that it has become the first company in Japan to deploy Nokia's MantaRay AutoPilot system, an AI-driven system that automatically optimizes mobile network quality, effective June 19.*1 To accelerate deployment, DOCOMO implemented MantaRay AutoPilot on a public cloud, marking the world's first case of optimizing a commercial mobile network via this system on a public cloud.*2

DOCOMO previously introduced Nokia's MantaRay SON*3 in November 2025, reducing manual workloads while enabling faster and more precise network-quality improvements.*4 Although MantaRay SON automated network optimization through closed-loop control—continuously monitoring network conditions and automatically adjusting settings accordingly—pre-design of parameters and configuration policies still had to be performed manually, making it challenging to configure parameter design and policies in real-time based on changing congestion patterns.

MantaRay AutoPilot eliminates this requirement by automatically designing network parameters and policies, and executing real-time optimization. With MantaRay AutoPilot, DOCOMO simply specifies desired quality targets, known as “intents,”*5 and the AI then analyzes base station quality and performance to determine the optimal parameter design and execution schedules. Working seamlessly with MantaRay SON, the system can complete a full optimization cycle in as little as 15 minutes.

By continuously repeating this process around the clock, the AI dynamically optimizes parameter design based on congestion patterns across locations and times of day, a task that was difficult for MantaRay SON alone. This enables ongoing network optimization and more stable data communication across a wide range of usage environments.

To implement MantaRay AutoPilot, DOCOMO adopted a public cloud architecture that optimizes the network directly from the cloud. This approach enabled rapid implementation without being constrained by hardware procurement lead times. DOCOMO also plans to integrate the system with various cloud-based AI platforms in the future.

Through its deployment of MantaRay AutoPilot, DOCOMO intends to achieve Autonomous Networks Level 4, the level of network autonomy defined by the global industry association TM Forum.*6 At this level, AI can predict network conditions and autonomously manage the network without human intervention.

DOCOMO will continue evaluating the system's effectiveness in commercial operations, aiming to advance AI-driven network management to deliver higher-quality communication services to customers.

  1. Based on DOCOMO research as of June 22, 2026.
  2. Based on DOCOMO research as of June 22, 2026.
  3. A self-organizing network (SON) automatically optimizes radio and network conditions through base station collaboration. Nokia's MantaRay SON is one such product.
  4. From the Nokia press release (Nov. 25, 2025): “ Nokia introduces MantaRay SON to NTT DOCOMO's multi-vendor 5G network ” (New windowhttps://www.nokia.com/newsroom/nokia-introduces-mantaray-son-to-ntt-docomos-multi-vendor-5g-network/).
  5. An operator's targets for its system (e.g., maintain communication quality in a specific area above X Mbps).
  6. A global nonprofit alliance of telecom and IT companies that establishes industry standards and best practices.

All company, product, and service names are trademarks or registered trademarks of their respective owners.

Appendix

Technical Overview of MantaRay AutoPilot

  • 1. Overview
    MantaRay AutoPilot is an AI-driven network quality optimization system developed by Nokia. Working in tandem with MantaRay SON, the optimization platform previously deployed in DOCOMO's mobile network, it enables autonomous network operations based on “intents” (desired quality targets). This represents a significant shift from conventional rule-based automation to continuous, sophisticated AI-driven autonomous optimization.
  • 2. System Architecture and Operation Process
    The system consists of three core components: base stations, MantaRay SON, and MantaRay AutoPilot. It autonomously optimizes the network through a continuous four-step cycle:
    System Architecture and Operation Process image
    1. Data Collection: Real-time performance data, including communication quality and traffic volume, is collected from base stations managed by MantaRay SON.
    2. AI Analysis and Decision-Making: Using the collected performance data and operator-defined intents (e.g., maintain communication speeds in a specific area above XX Mbps), MantaRay AutoPilot's AI analyzes, learns, and evaluates the network status. It then autonomously determines which parameters require improvement and the optimal timing for execution.
    3. Optimization Directives: Based on the AI's evaluations, quality optimization directives are sent to MantaRay SON at intervals as short as 15 minutes.
    4. Optimization Execution: Upon receiving these directives, MantaRay SON remotely reconfigures the target base stations to execute the network quality optimization.
  • 3. Comparison with Conventional Operations

    In previous operations that relied solely on MantaRay SON, the optimization cycle ran daily. Furthermore, analyzing which configuration changes to apply and measuring their post-implementation effectiveness required manual intervention.

    MantaRay AutoPilot uses AI to fully automate the entire optimization cycle, from parameter analysis and configuration to post-implementation performance evaluation. This reduces the cycle time to as little as 15 minutes, enabling faster, more granular network-quality improvements without manual intervention.

Go to top of the page