Start with an inventory of existing and planned camera positions. Record connection settings, assigned tasks, network links, compute nodes and site map areas. Save the network, camera, calibration and map configuration. Choose a representative group of new cameras for a trial. The proposed target scale still needs on-site testing.
Add cameras in groups and measure load
For each group, record connection results, peak bandwidth, resource use, processing delay, queue backlogs and reconnection. If work will be split across compute nodes, check clock sources and matching records while the nodes exchange results. Collect evidence during existing operations, activation of new tasks and fault recovery.
Check the site map and overlap areas
Mark the added positions on the same site map. Use site control points to check affected mappings at new viewpoints, existing handoff zones and new occlusions. Record the coordinate reference, calibration version and tested areas.
Check recovery and rollback
Agree on the sequence for disabling added cameras, restoring resource allocation and restoring calibration. Test restoration of existing video after rolling back a problematic batch. Evaluate expansion acceptance using operating records at the target scale.
Related service: Multi-camera synchronization, calibration and fusion.