Boston Dynamics released Spot 5.2, a major software update that positions Spot as a bridge between enterprise AI systems and physical inspection operations. The release redesigns Orbit fleet management software around industrial assets rather than individual inspection tasks, allowing users to import asset hierarchies from CMMS, MES, and EAM platforms. Spot can now be dispatched autonomously by AI agents when predefined conditions are met — for example, a fixed sensor detecting a temperature anomaly can trigger Spot to inspect the affected equipment without human intervention.

The update expands visual inspection capabilities using Google Gemini. AIVI-Learning now analyzes video to detect conditions difficult to spot in still images, including water dripping from pumps, slipping conveyor belts, and equipment warning lights. Site Scans enable Spot to continuously monitor facilities for temporary risks like spills, unauthorized personnel, or unattended objects — broadening its role from scheduled equipment checks to real-time environmental observation.

New predictive maintenance tools include gas sensing for 20+ gas types, visual vibration analysis using Spot Cam to amplify imperceptible machinery movement, and acoustic partial discharge monitoring for high-voltage equipment. These capabilities stream data directly into Orbit alongside standard inspection results, providing operators with a unified view of facility conditions.

The release reflects a shift toward AI-orchestrated physical operations. By opening Orbit's API to external data sources and implementing a Model Context Protocol layer, Boston Dynamics enables workflows where enterprise software can autonomously direct robotic action. A CCTV system detecting movement can dispatch Spot to investigate. A predictive maintenance platform can instruct Spot to inspect equipment only while it's operating. The result is a system where AI-generated insights lead directly to autonomous physical interventions without manual tasking.