Boston Dynamics builds no headsets. Its robots are the mobile sensors that keep industrial digital twins current, trained first in simulation and sometimes piloted through VR.
Boston Dynamics Prime
Boston Dynamics avoids consumer "metaverse" positioning entirely. Its strategy centers on embodied AI, industrial automation and providing the physical data layer needed to create living digital twins of real environments. Spatial computing is the bridge for high-fidelity reality capture, remote robotic teleoperation and physics-accurate simulation. The company builds no XR headsets; it equips its robots (Spot, Stretch and Atlas) as mobile sensors and uses third-party platforms, mainly NVIDIA Omniverse, Isaac Sim and advanced VR interfaces, for visualization, simulation training and remote control.
Metaverse Products
- Orbit robot fleet management: The cloud-based command center that orchestrates robot fleets, schedules autonomous missions and processes site data. Its Site View feature aggregates 360° imagery and spatial data over time, so users can walk virtually through a historical visual record of their facility.
- Spot with reality capture payloads: The quadruped robot as an autonomous mobile scanner. With payloads such as Spot CAM+IR, Leica BLK ARC LiDAR or Trimble scanners, it crosses rugged terrain to produce high-accuracy 3D point clouds.
- NVIDIA Omniverse and Isaac Lab integration: The simulation infrastructure where AI policies are trained at scale through reinforcement and imitation learning before code reaches physical robots.
- Stretch warehouse automation software: The software behind the Stretch truck-unloading robot, using 3D vision to perceive, categorize and model stacked boxes in real time.
Metaverse Applications
- Dynamic industrial digital twins: Spot patrols and updates 3D replicas of factories, power plants and construction sites, feeding twin platforms for remote monitoring, deviation detection against CAD and BIM models, and predictive maintenance.
- Immersive robotic telepresence: In research and advanced pilots, operators use VR headsets and controllers to "embody" Atlas or Spot for precise two-handed tasks in hazardous or remote places.
- Synthetic training and policy simulation: Millions of parallel simulations in Isaac Lab and Omniverse accelerate locomotion, dexterity and task behaviors, so robots "learn" to walk and grasp digitally.
- Automated construction documentation: Daily 3D captures flow into Building Information Modeling workflows to check as-built conditions against plans.
Examples of Boston Dynamics Product Enabled Use Cases
- Facility twin maintenance (Spot with Leica BLK ARC): Autonomous scanning updates factory twins in real time, replacing manual surveys and revealing wear or structural change.
- VR-guided manipulation (Atlas): Engineers in VR guide the electric Atlas through precise tasks, such as moving automotive parts, to gather data for imitation learning.
- Remote inspection through Orbit Site View: Plant managers walk through 360° panoramas captured by Spot to diagnose tripped breakers or safety hazards from anywhere.
- AI training in simulation: Reinforcement learning in Isaac Lab tests millions of locomotion scenarios across virtual terrain before deployment.
- BIM compliance on construction sites: Spot's daily scans overlay on architectural models to flag deviations early, cutting rework and delays.
Presented by Hkdolts at the Metaverse Standards Forum Use Case Task Force Spotlight Series, 3 June 2026.