Summary

  • DOCOMO and NTT DOCOMO BUSINESS agreed with Starlink to use Starlink Mobile V2 for planned voice, faster internet browsing and direct IoT communications during fiscal 2028.
  • The announcement does not yet define capacity, latency, traffic priority, certified devices, pricing, enterprise service levels or fallback behaviour.

A contract changes the probability of a roadmap. It does not yet define the service. DOCOMO's 14 August announcement moves direct-to-device satellite communications beyond a messaging supplement: voice calls, faster internet browsing and IoT connectivity are planned during fiscal 2028. Those functions could make the satellite layer relevant to everyday continuity, remote operations and enterprise monitoring. The conditions for each function, however, are explicitly left for later.

The current baseline matters. docomo Starlink Direct began on 27 April with messaging and selected application data on compatible phones, outdoors in terrestrial and Wi-Fi gaps with an unobstructed sky. It is free for the time being and ordinary voice calling is not supported. DOCOMO says 6.5 million compatible-device customers had connected by the end of July. That demonstrates distribution; it does not supply a performance baseline for the V2 functions.

Voice is the sharpest change because it creates an expectation of continuity rather than delayed delivery. A useful service definition would specify call setup success, median and tail setup time, audio quality, drop rate, session duration and behaviour when a handset moves between terrestrial and satellite coverage. It would also state whether emergency and public-priority calls can pre-empt ordinary traffic, and what the ordinary customer sees when that happens.

“High-speed internet” needs an equally concrete envelope. The phrase is not a throughput distribution, a concurrency model or an application guarantee. DOCOMO should define expected downlink and uplink ranges, latency under ordinary and congested conditions, traffic management, session limits and the share of satellite capacity available to each service class. A best-case demonstration would not establish the experience of many people browsing in the same beam.

IoT widens the control surface further. A field sensor is not a smartphone with a smaller screen. Certification must cover antennas, power budget, duty cycle, firmware update, authentication, data integrity, roaming and failure recovery. Remote monitoring in mountains, islands and at sea can create long-lived operational dependence. Enterprises need to know whether an alarm is delivered, how late data may arrive, how a device proves identity and what happens when the satellite layer is unavailable for hours.

The corporate route through NTT DOCOMO BUSINESS makes service levels more than a technical detail. An enterprise offer needs a support boundary, incident notification, availability definition, restoration objective, maintenance treatment and credits or remedies. It also needs a clear division of responsibility between the customer, DOCOMO, its business arm and the satellite supplier. Without that map, a failed remote sensor or voice session can become an unresolved chain of vendors.

The FY2028 date should therefore be treated as a sequence of verifiable gates. Spectrum and regulatory approvals come first, followed by handset and IoT certification, integration with the terrestrial core, laboratory load tests, field trials, priority exercises, commercial terms and a public launch definition. Each gate can pass at a different time. Publishing the sequence would make a delayed feature understandable without turning the whole programme into a binary success or failure.

Supplier concentration is the strategic counterweight. A broader satellite service can reduce dependence on terrestrial towers in remote areas while increasing dependence on one orbital platform, its capacity allocation and its upgrade schedule. DOCOMO can preserve control through portable device standards, operational telemetry, audit rights, traffic-priority rules, incident data and an exit or substitution plan. Those are not objections to the agreement; they are the terms under which it becomes infrastructure.

The agreement is thus more important than another capability announcement. It creates a route from occasional messaging to a network layer that may carry conversations, browsing sessions and machine telemetry. The next decisive document is not a promotional countdown to fiscal 2028. It is a service envelope that tells customers what can be attempted, what is expected to complete, which traffic wins under pressure and how the system fails safely.

Sources