The live production of international sporting events is undergoing fundamental technical modernization. Ahead of the IBC broadcast convention, technology providers presented dedicated systems that integrate artificial intelligence directly into professional television signal paths. These developments follow two clear trajectories: automated localization for global audiences and real-time visual data processing at the edge of the pitch. Both the audio specialist AI-Media and chipmaker NVIDIA showcased corresponding solutions designed for broadcast facilities and outside broadcast vans in early September 2026.
AI-Media, listed on the Australian Securities Exchange, introduced two specialized hardware appliances on September 3, 2026: the LEXI Text Encoder and the LEXI Voice Encoder. The new hardware units were engineered for integration into professional broadcast environments and support legacy SDI and IP connections as well as the SMPTE ST 2110 standard. This open network standard forms the backbone of modern, uncompressed IP-based studio architectures. By offering native integration, the devices allow broadcasters to adopt automated processing pipelines without requiring complex external format conversions.
From a functional perspective, the two LEXI systems tackle a persistent bottleneck in international sports coverage. The system ingests original live commentary and converts it into accurate subtitles across multiple languages with minimal latency. In addition, the algorithms generate synthetic voice tracks and accessible audio descriptions for visually impaired viewers. To prevent transcription mistakes during fast-moving athletic sequences, the underlying models were specifically trained on sports terminology and athlete rosters.
NVIDIA pursued a complementary approach focused on visual computing with its Sports Intelligence framework, announced in early September 2026 around IBC. Developed in collaboration with Dell and HP, the solution runs on dedicated Blackwell workstations. The software framework relies on NVIDIA Holoscan for Media, an architecture purpose-built for low-latency media processing in production environments. This configuration allows broadcast teams to deploy heavy computational models directly on site or inside mobile control trucks.
The core capability of NVIDIA's platform centers on real-time computer vision during live competition. Multimodal AI models ingest direct camera feeds to calculate precise spatial coordinates for the ball and every athlete on the field simultaneously. By matching these tracking metrics against historical match databases, the framework generates situational performance metrics and visual graphics on the fly. Director teams can instantly incorporate these contextual insights into the broadcast feed without relying on delayed third-party data providers.
Together, the announcements from AI-Media and NVIDIA illustrate how deeply edge artificial intelligence is penetrating the broadcast pipeline. Combining low-latency audio localization with automated visual analytics allows rights holders to serve global viewers with personalized and accessible programming. AI-Media eliminates language barriers on the audio side, while NVIDIA delivers the visual depth required for next-generation sports presentations. For international broadcasters, this convergence represents a major shift toward automated, data-rich live television workflows.