As Indosat expands its AI-RAN infrastructure, telecom leaders gain a practical blueprint for improving network efficiency and delivering new services.
Telecommunications executives are under constant pressure to increase returns on recent cellular investments. Consolidating hardware and running network functions alongside AI workloads on shared compute offers a pathway to lower capital expenditure and enable innovative service models.
Following a commitment announced at MWC with partners Nokia and NVIDIA, Indonesian operator Indosat is moving beyond lab prototypes. The company is building an AI-powered grid intended to serve both dense urban centres and more remote rural communities.
Validating AI-RAN hardware for network efficiency
Combining radio access processing and AI applications on shared GPUs requires careful planning and operational changes. Indosat validated this shared-hardware approach by conducting the first AI-driven 5G call in Southeast Asia.
In a live demonstration, the operators showed ultra-low latency capabilities by remotely controlling a robot in Surabaya in real time.
Replacing legacy, purpose-built network appliances with flexible compute clusters also requires new processes and local skills. To close this gap, Indosat opened a dedicated facility in Surabaya where students from leading local universities will collaborate with the operator to build use cases across healthcare, education, and agriculture.
Vikram Sinha, CEO of Indosat, frames this social impact as a core business priority. “When you talk about AI-RAN, I think it is much more than technology,” he said. “We are working hand-in-hand on making sure that what we do helps people.”
Building collaborative infrastructure models
Teams evaluating private networks or modern RAN approaches can view this AI-RAN deployment as an example of effective cross-industry collaboration. Scaling advanced network architectures requires breaking down internal silos and engaging with a broader ecosystem of partners.
An industry consortium is already taking shape, including Nokia, NVIDIA, Indosat, SoftBank, and T-Mobile US. Such collaborations let participants pool expertise, share implementation lessons, and spread development costs.
Sinha urges peers to commit fully once they choose this path. “First thing first, you have to believe in it,” he said. “And if you believe in something, go all in.”
The cooperative approach is proving viable despite early skepticism from some observers. Technologies like AI-RAN deliver real value only when they translate into operational efficiencies and solve tangible problems for network users and enterprises.
Related coverage: Qualcomm: preparing telecom infrastructure for AI-native 6G rollouts.
Interested in AI and big data insights from industry leaders? The AI & Big Data Expo takes place in Amsterdam, California, and London. The event is part of TechEx and is co-located with IoT Tech Expo and Cyber Security & Cloud Expo.
Telecoms coverage is provided by TechForge Media. Explore other enterprise technology events and webinars organized by the publisher.