Prompt-based AI camera control, defined
Prompt-based AI camera control is the use of natural language instructions and scene-aware algorithms to direct robotic PTZ cameras in real time, allowing systems to identify subjects, follow action, and trigger framing or production cues without continuous manual operation. At InfoComm, PTZOptics and LayerJot are turning that definition into a working live production automation demo. Their setup links PTZOptics SimplTrack3 robotic PTZ cameras with LayerJot’s agentic AI software and Intel Core Ultra Series 3 processors, showing how on-device AI camera control can identify, follow, and respond to objects and events as they happen. Instead of relying on cloud services or large control rooms, the system runs locally, aiming to give broadcasters, educators, and pro AV teams a prompt-based AI broadcasting toolkit that fits into existing rooms, lecture halls, and studios while cutting down on repetitive camera operation.
Inside the PTZOptics–LayerJot live production workflow
At the heart of the collaboration is PTZOptics’ Visual Reasoning initiative, which turns video from robotic PTZ cameras into actionable data for camera moves and external triggers. LayerJot’s Agentic Visual Application connects directly to the SimplTrack3 camera, applying scene understanding and AI prompting to decide when to track a speaker, zoom to a key area, or reframe a shot. Intel Core Ultra Series 3 processors provide the local compute needed to keep these decisions in sync with live action. According to Intel’s Dr. Ben Cope, these processors “deliver the on-device AI compute performance that makes local visual reasoning practical at scale.” For production teams, this means fewer manual joystick moves and more consistent automated framing, while still allowing operators to step in when creative or editorial choices are needed during a multi-camera broadcast.
From manual ops to autonomous live production automation
The PTZOptics and LayerJot demo points toward a future where live production automation is defined less by fixed rules and more by intent expressed as prompts. Instead of programming complex macros, operators can issue natural language instructions such as following a presenter, prioritizing a panelist, or highlighting a section of the stage. LayerJot’s scene understanding then translates those prompts into camera control actions across the SimplTrack3 robotic PTZ cameras. PTZOptics describes this approach as enabling systems to “describe, count, understand and trigger actions based on what cameras see,” with applications in broadcast, pro AV, education, healthcare, manufacturing, worship, and live events. For smaller teams, the appeal is clear: one operator can supervise AI camera control across several angles, shifting from basic device management toward higher-level creative and editorial decisions during live streaming.
InfoComm as a launchpad for prompt-based AI broadcasting
InfoComm’s showfloor context matters: vendors are not only showing individual products, but whole workflows that merge AV, IT, and media. Diversified’s “The Pitch” installation, for example, combines live content, visualization, audience engagement, and production tools into one connected experience, reflecting a broader demand for integrated platforms instead of isolated devices. In the same North Hall, the PTZOptics and LayerJot demo positions AI camera control as a core element in this shift, turning cameras into active participants in the workflow rather than passive endpoints. As exhibitors move away from one-off feature demos toward complete, scalable deployments, prompt-based AI broadcasting becomes a practical path to simplify multi-camera operations. InfoComm, with its focus on immersive experiences and data-driven environments, is emerging as the proving ground for autonomous live streaming setups that can be deployed on-site without large crews or cloud-heavy infrastructure.
