Scaling Oriented Imagery to 500K+ panoramas

[chinmay-terrahelix]

Hi all,

We're evaluating oriented imagery for a large infrastructure project that would involve around 500,000 360° panoramas, with potential to grow to 1M. Images would sit on AWS S3, published as an Oriented Imagery Layer to AGOL, and consumed via the OI Viewer widget in Experience Builder.

We haven't been able to find documented performance guidance for catalogues at this scale, so a few questions:

- Has anyone deployed an OIL with 500K+ exposure points? What was the experience like?
- Should we be splitting into multiple regional layers, or can a single OIL handle this volume if properly provisioned?
- Is AGOL realistic for this, or should we be looking at Enterprise with Image Server?
- Any recommended configurations for the underlying feature service (maxRecordCount, spatial indexing, etc.) to keep the widget responsive?

Any pointers would be very helpful at this stage.

Thanks,
Chinmay

[ClarkSelby]

Chinmay, one approach you might want to consider is using tools like Vidgeo and Site Viewer 360 alongside (or in place of) the standard OI Viewer, depending on your workflow. These are designed for large-scale 360° video and imagery in ArcGIS and can reduce the need to manage massive collections of discrete panoramas.

For example:

• With Vidgeo, you can stream 360° video and dynamically derive viewpoints rather than managing hundreds of thousands (or millions) of individual images. More info is available here: https://www.nodeology.net/vidgeo/

• Site Viewer 360 is optimized for navigating large volumes of 360° content and helps keep the user experience responsive at scale. More info is available here: https://www.nodeology.net/site-viewer-360/

Both approaches can leverage AWS S3 for storage, while enabling a more efficient, streaming-based architecture.