Hosted by the Global Satellite Operators Association (GSOA), the webinar “Empowering Autonomous Satellite Networks: AI, ML, and the Cloud Revolution” brought together industry leaders from Microsoft Azure, Intelsat, ST Engineering, and Speedcast to discuss how new technologies are transforming the future of satellite communications. As global connectivity demands rise, AI, machine learning, and cloud-native infrastructure are enabling a new era of intelligent, autonomous satellite networks.
The panel of speakers included:
- Sridhar Kuppanna, CTO and Senior Vice President of Engineering at ST Engineering iDirect
- Minh Nguyen, Architect on the Emerging Technologies Team at Microsoft Azure
- Will Mudge, Executive Vice President and CTO at Speedcast
- Ken Takagi, Director of Business Development at Intelsat
ST Engineering: Building Resilient, Intelligent Networks
Kuppanna emphasized that AI is essential in the Satcom industry and is no longer experimental, stating, “The technology is here… it has demonstrated that it’s mature and reliable.” He explained that AI is foundational to solving real-world Satcom challenges and shared his team’s goal to help propel the industry toward more autonomous networks and improved operational efficiency. Key areas of focus include self-healing applications, which allow the network to automatically detect and resolve issues, as well as digital twin technology, which can simulate real-world conditions to support smarter decision-making and predictive operations.
Microsoft Azure: Trustworthy AI and Autonomous Tasks
Nguyen discussed how Microsoft Azure is helping customers build trusted AI systems for vital space missions, including Earth observation and satellite communications. A key challenge, he explained, is confidence in AI decision-making. “AI in general is non-deterministic… so how do we trust this technology to make decisions on our behalf?” He highlighted Microsoft’s work in agentic orchestration, where users issue natural language prompts and AI systems autonomously handle satellite tasking and data collection. This method accelerates actionable insights while reducing operator workload in Earth observation. Nguyen also mentioned efforts in multimodal AI systems and small, edge-deployable models that bring intelligent processing closer to the point of data generation.
Speedcast: Predictive Maintenance at the EDGE
Mudge shared that Speedcast began using AI several years ago with predictive analytics designed to identify and resolve problems before they occur. “We started to switch over to predictive data to track things that are going to break before they break,” he said. To enable more intelligent decision-making at the edge, Speedcast is combining data from across its networks such as weather, signal strength, and latency measurements. These capabilities, he noted, are helping the company modernize its support infrastructure and better meet the evolving needs of its customers.
Intelsat: Smarter Networks with AI, 5G NTN, and Ubiquitous Connectivity
Takagi explained how Intelsat is undergoing a major transition from a legacy hardware-first provider to a flexible, software-defined network powered by AI. He said that the rollout of 5G NTN, rising customer expectations, and the demand for ubiquitous connectivity are what are driving this evolution. He pointed out the combined pressures of market disruption from LEO players and the need to support high-speed, high-reliability service across geographies and verticals.
To meet these demands, Intelsat is integrating AI into its backend and customer-facing operations. Takagi highlighted their AI-powered customer portal, which enables customers to monitor network performance, activate terminals, and provision services in over 80 languages. He also emphasized efforts to use AI for predictive analytics, particularly to predict network congestion and dynamically allocate bandwidth in real time. These are especially critical in mobility sectors like aviation and maritime, where flexibility, low latency, and continuous coverage are essential.
The Future of Autonomous Satcom
A shared theme across all the panelists was that cross-industry collaboration, data quality, and trust in AI, are essential to enabling resilient, autonomous, satellite networks.
