Questions Regarding Shared Spatial Anchors Limits and Large-Scale Deployments

Hello,

We are evaluating the feasibility of using Meta's Colocation / Shared Spatial Anchors SDK for a large-scale VR/MR project and would appreciate some clarification regarding its scalability.

While reviewing the available documentation, we found limited information about supported participant counts, with most examples referring only to "two or more" users.

Our current prototype is intended to scale to approximately 200 concurrent users operating within a relatively large shared physical environment. To assess whether Shared Spatial Anchors are suitable for this use case, we would like to better understand any practical or hard limitations of the system.

Specifically, we are interested in the following:

A. Concurrent headset support

* Are there any documented or recommended limits on the number of headsets that can share the same spatial anchor data?
* In our scenario, a single headset (or a small number of headsets) could act as the source of the spatial anchors, while all other headsets only receive and localize against those anchors.
* Is there a maximum number of concurrent devices that can reliably receive and use spatial anchor data from a single source?

B. Physical environment size

* Are there any limitations regarding the size of the tracked environment when using Shared Spatial Anchors?
* For example, can the system reliably support a shared space of approximately 200 m², or are there recommended maximum dimensions/areas for accurate localization and alignment?
* If limits exist, are they determined by the anchor system itself, point-cloud tracking capabilities, network considerations, or another factor?

Additionally, if there are any best-practice guidelines, performance benchmarks, or case studies involving deployments with dozens or hundreds of simultaneous users, we would be very interested in reviewing them.