Can Human Operators Keep Pace with Next-Gen Constellations?
Jun 08, 2026
At the ATxSG 2026 conference in Singapore, Sridhar Kuppanna, Chief Executive Officer of ST Engineering iDirect, outlined his vision for the next generation of intelligent satellite networks, emphasizing that the industry has moved well beyond the experimentation phase and into an era of real, measurable AI outcomes.
ST Engineering Manufacturing Competence Center. Credit: ST Engineering iDirect
Kuppanna said that enterprises are no longer experimenting with AI in a lab setting, but are already using it in actual operations. For satellite and hybrid network operators, three capabilities are now a must, he said. The first is domain-specific AI, which involves models that are trained on satellite network data instead of general data sets. The second is closed-loop automation, which enables systems to predict, decide, and act without human intervention. The third is a single data platform that can normalize the stream of telemetry data in multi-orbit and multi-vendor environments. His perspective is that the future of AI is not about scaling up models, but about integrating intelligence into the network fabric, enabling it to learn, optimize, and troubleshoot on the fly.
One of the key topics he discussed was the integration of satellite, terrestrial, cloud and edge infrastructure. The 5G Non-Terrestrial Networks standard is the key driver here, providing a common interoperability framework between space and terrestrial systems. Cloud-native architectures are driving elastic scaling and distributed intelligence, and AI at the network level is making real-time decision-making possible in challenging applications like maritime operations, defense, and remote industrial sites. The convergence is changing the nature of partnerships in a fundamental way. Satellite operators are increasingly collaborating with cloud hyperscalers to provide distributed computing and AI model hosting, with mobile network operators to provide seamless hybrid connectivity, with device and terminal manufacturers to introduce NTN-ready intelligence at the edge, and with AI platform providers to speed up automation and orchestration. The companies that will spearhead this market will be creating a single, standards-based, multi-access experience, not isolated proprietary networks.
Three technologies are most promising for the future of the satellite and global connectivity industry. Perhaps the biggest development is agentic AI. Agentic AI is capable of reasoning, planning, and executing multi-step tasks in large network environments with minimal human supervision, compared to traditional AI that works on a single task at a time. In parallel, autonomous network architectures that integrate AI processing, real-time telemetry collection, and closed-loop automation are facilitating self-healing infrastructure, which is able to detect faults, diagnose root causes, and restore performance without requiring a technician to intervene. The third priority is quantum-safe security. With the growing complexity of satellite constellations, and multi-orbit operations becoming the norm, the protection of communications infrastructure against next-generation threats is no longer a future problem but a current engineering requirement.
Kuppanna concluded with a broader observation about the trajectory of the industry. As networks become more dynamic, more distributed, and more dependent on real-time responsiveness, human-driven operations will simply be unable to keep pace. The scale and complexity of managing multi-orbit satellite systems, hybrid terrestrial connections, and cloud-based services simultaneously demands a level of speed and precision that only autonomous, AI-powered infrastructure can reliably deliver. His message to operators and technology partners underscored that the next decade in connectivity belongs to intelligent networks that can sense their environment, adapt to changing conditions, and maintain predictable performance anywhere on the planet, making now the time to build that foundation.
