Intelligence for your cameras
Find out if your cameras
can do more.
Tell us what equipment you have. Get initial compatibility guidance and a next step, with no charge for this enquiry.
Initial enquiry at no charge. Installation and service are quoted separately.
Let’s start with your equipment
Discover what your
cameras could do.
Tell us what you have and what you want to detect. We’ll guide you on compatibility and the next step toward a quote.
One request, one conversation.
We prepare a message with your choices. You review and send it through WhatsApp.
Test at a real business
Existing cameras.
Measured performance.
Real footage. Measured capacity.
Cloud pilot · (UTC)
- Cameras processed
- 2Business interior and terrace.
- Analysis per camera
- 5.0–5.2 FPSFrames analyzed per second.
- Skipped FPS (maximum)
- 0Reported across the 60 samples.
- Server CPU (peak)
- 53.3%Total usage, including other services.
60 samples over 15 minutes. Historical snapshot; no connection to the live system.
Measure what my cameras can doHow we measured it and what it shows
The pilot processed two feeds at 704 × 480, targeting 5 FPS per camera, on a cloud server with 4 assigned CPUs and a 2500 MiB container RAM limit. The detector was OpenVINO MobileNet on CPU; primary recording stayed on site. Available server RAM reached a minimum of 5356 MiB; this is not detector memory usage.
The saved summary reported 5.0–5.2 FPS for the interior and 5.0–5.1 FPS for the terrace. Zero skipped FPS refers to sampled polls, not a continuous-service guarantee. Analysis FPS differs from recording FPS.
The video and metrics come from separate tests. Replay boxes use saved observations from a different detector, YOLOv9-t / ByteTrack. This measurement establishes processing capacity during that window; detection accuracy, identity, arrivals and counts require separate testing.
View the metrics summary and provenance (JSON)