.png){kind=logo}
Newsletter N°91 - Summer Edition 2026
📶 Telecom: Rakuten achieves Level 4 autonomous network
Rakuten Mobile has achieved the world's first TM Forum Level 4 autonomy validation for RAN energy-efficiency optimisation in a live commercial Open RAN network. The certification, obtained in February 2026, recognises Rakuten's ability to use AI-driven, closed-loop automation to optimise network operations with minimal human intervention.
Rakuten Mobile 5G Sub6 base station, part of the operator’s Open RAN infrastructure. (Photo: Rakuten Mobile, Inc.)
At the core of the solution is Rakuten's cloud-native Open RAN architecture, combined with the Rakuten Symphony platform and its RAN Intelligent Controller (RIC). AI and machine-learning applications continuously analyse network KPIs, identify traffic patterns and forecast future demand. Based on these insights, the system can automatically adjust network resources and energy consumption in real time.
This approach enables Rakuten to reduce RAN electricity consumption by approximately 20% compared with its existing network, while maintaining the required quality of service. Rather than relying on predefined rules, the system operates through a closed feedback loop: it observes network conditions, makes an optimisation decision, implements it and continuously evaluates the result.
The Level 4 certification is significant because it represents a step beyond conventional network automation. TM Forum defines Level 4 as a stage where networks can adapt autonomously to unexpected situations, using AI and machine learning to make decisions with limited human oversight. At lower levels, automation generally remains constrained by predefined workflows and scenarios; at Level 4, the network can independently determine how to respond to changing conditions.
Rakuten argues that its Open RAN architecture is particularly suited to this evolution because its open, software-driven and cloud-native design allows AI to access and act on network data more flexibly than traditional, siloed architectures. The company has also developed its own software, the Rakuten Intelligent Controller, to detect and predict network issues and optimise operations automatically.
The initiative is part of a broader shift towards self-managing telecom infrastructure, where AI moves from being an analytical support tool to becoming an operational layer capable of making and executing network decisions. Rakuten is already looking beyond energy optimisation, with the objective of extending Level 4 capabilities across the network from the core to the edge and ultimately moving towards Level 5 fully autonomous operations.
For telecom operators, the development illustrates how AI + Open RAN can simultaneously address operational efficiency and sustainability. By allowing networks to dynamically adapt to traffic demand, operators can reduce unnecessary energy consumption while moving towards networks that require progressively less manual intervention.