Boundaryless and contextual-boundaryless safety best practices

• Combining Scene API with Depth API supports proper depth order between virtual content and the physical space. If not used, virtual content will block the user’s view of physical objects.
• Place head-locked content to the side of the field of view so that it does not block potential trip hazards or obstacles in front of the user (in other words, avoid the user’s lower and central field of view).
• Reduce virtual content opacity for better user visibility when moving at high speeds, moving in reverse, or getting close to physical objects or room walls, especially if the content is not intended for close interaction.

• Always turn on the boundary before transitioning from passthrough primary mixed reality to blended or fully immersive experiences.
• When transitioning between mixed reality and fully immersive experiences, provide time for users to reorient and adapt to changes before introducing task demands.
• Consider the amount of virtual content included relative to the type of activity required of the user. For experiences with a high degree of movement, consider decreasing the amount and size of virtual content that could obscure the user’s view of their physical activity space.

• Placing large virtual content on the floor that blocks clutter or trip hazards from view.
• Having many virtual objects which can cumulatively block the user’s view of their physical space. This includes high levels of passthrough stylization and virtual reskinning of the scene.
• Placing virtual objects or characters beyond a physical wall, or inside phyiscal world objects (this creates improper occlusion which causes misperceptions in depth and actionable space).
• Placing large virtual objects directly in front of physical objects and walls, especially those intended for interaction where the user gets close.
• Experiences that require the user to back-up quickly, dodge, run, or flee, as this increases the possibility of unintentional trips, contacts, or breakage.

• Avoid encouraging users to move outside of the boundary. Space outside the boundary may contain clutter or obstacles and because the user has a reduced field of view, these obstacles will be harder to detect.
• Dynamic elements (for example, pets and people) are not part of the scene, and can be hidden from view by virtual content. Depth API can support both static and dynamic objects and update occlusions based on user movement and therefore it is recommended that you use both simultaneously.
• Scene is not a safety feature. It does not provide safety warnings when users get close to physical objects. The MR Utility Kit has a wall boundary feature that alerts users when they are close to room boundaries, but keep in mind it does not alert them to objects within the defined activity space.
• Display resolution and color passthrough on Meta Quest 3 improves depth perception, visual acuity, and object recognition compared to Quest 2.
• Verify Boundary API implementation in headset over PC/Link to ensure proper boundary behavior.