“We only learn a camera was useless when we go to pull footage.”
VMS dashboards say the recorder is up. They don't say the view drifted, the lens fogged, or the recording has a 4-hour gap.
CCTV fleet readiness · on-prem
Camera health monitoring that runs where your cameras do.
CHEQIT watches every camera across every site for the things VMS uptime won't tell you — blur, freeze, darkness, scene drift, recording gaps — and turns the failures into a ranked repair queue for security and IT teams.
Built for businesses, enterprises, and local government teams responsible for cameras across depots, campuses, facilities, streets, and remote sites.
On-prem
No video egress
Offline activation
Scales by camera
Cancel anytime
Why teams buy CHEQIT
Large CCTV estates fail quietly. CHEQIT makes failures visible and actionable.
“Manually checking thousands of streams isn't a job.”
Spot-checking is unreliable at any scale. Engineers find the issues that have already cost evidence — not the ones forming now.
“Cloud-only doesn't fit our network.”
Restricted, regulated, or air-gapped sites can't ship video to a vendor cloud. Monitoring has to live inside the perimeter.
What CHEQIT checks
A degradation taxonomy purpose-built for CCTV — not a generic uptime probe.
CHEQIT ships with detectors across availability, image quality, stream behaviour, network, encoding, security posture, and hardware/optics. They run on a tiered pipeline tuned for fleets up to 10,000 cameras with bounded compute.
Selected camera-view examples
1 / 11
Example: Blur
Soft focus, motion smear, or lens contamination can leave a camera online but unusable for evidence review.
Supported alert families
Availability & continuity
VMS + network probesCamera offlineVMS monitor offline or endpoint unreachable, retry-confirmed.
VMS API timeoutPer-provider circuit breaker on repeated request failures.
Archive unavailablePlayback or recording-index retrieval fails.
Recording gapComputed from the recording index window, per camera.
Storage disk fullThreshold + forecast on recorder/host storage.
Retention driftFootage expiring outside the expected window.
Image & scene quality
Snapshot CV · evidence attachedBlur / defocusEdge-variance against per-camera baseline.
DarknessMean luma + dark-pixel ratio, time-of-day aware.
Overexposure / glareHighlight-clip ratio with bright-region hint.
Occlusion / paintEdge collapse + extreme luma; covered or sprayed.
Tilt / camera movedDrift versus baseline, with approved orientation support.
Scene changeSSIM/hash trend with accepted day/night modes.
Low contrast / fogContrast and entropy drop versus slot baseline.
Colour cast / IR-cut stuckDay/night transition errors and white-balance shift.
MacroblockingBlock-edge discontinuity from encoder/transport faults.
Stream & encoding behaviour
Stream sampling · metadata trendsCamera freezeStuck or near-zero diff over a confirmed window.
Duplicate-frame loopRepeating GOP/hash patterns the eye misses.
FPS driftFrame-rate falling below baseline, policy-aware per camera.
Resolution driftStream dimensions changed without a config update.
Timestamp driftFrame timestamps versus system NTP and VMS clock.
Bitrate driftSustained drop or surge versus stream baseline.
Codec changeH.264 ↔ H.265, profile / GOP / keyframe changes.
Flicker50/60 Hz luma oscillation; suppressed during night transitions.
Camera reboot loopWatchdog reboot cycles correlated with PoE events.
Network, security & hygiene
Hybrid · operational postureNetwork latencyRTT sampling with per-site baseline.
Jitter / packet lossBounded probe budget, correlated with camera-offline.
Tamper signalONVIF tamper events raise sampling cadence.
Default-credential riskHeuristic — never brute-force; admin-facing risk surface.
Firmware outdatedONVIF device info versus a curated allowlist.
Privacy mask unexpectedNew uniform overlays appearing without authorisation.
Autofocus huntingSustained sharpness oscillation — typically lens failure.
Banding / rolling linesFrequency analysis on stream samples.
Green / purple screenEncoder/GPU failures producing dominant chroma planes.
How it works
Connect your VMS, run on-prem, operate by exception.
Connect what you already run
Point CHEQIT at your VMS or directly at cameras over ONVIF/RTSP. No agents on cameras, no recorder changes.
Run inside your perimeter
A single signed Linux bundle, deployed by systemd next to the recorder. Activate online or with a signed key file for air-gapped sites.
Operate by exception
A ranked queue replaces dashboard scanning. Each alert ships with before/after evidence, archive video where the VMS provides it, downloadable incident clips, and controls for accepting valid baselines.
Built for distributed estates
For teams managing cameras across many buildings, locations, and network conditions.
High street CCTV · car parks · depots · social housing
Campuses · warehouses · branches · plants · DCs
Multi-tenant · per-customer scopes · service evidence
VMS-flexible architecture
Works with the video systems you already run.
CHEQIT connects to supported VMS platforms and direct camera sources, then presents camera health, evidence, and alerts in one consistent workflow. The monitoring experience stays the same as new integrations are added, so your VMS choice does not define what CHEQIT can detect.
Shinobi CEDirect integration, MVP-validated
LiveONVIF + RTSPDirect cameras, discovery + sampling
LiveAvigilon UnityCustomer-driven connector discovery and API validation are under way
In progressGenetec Security CenterFuture integration candidate
PlannedMilestone XProtectFuture integration candidate; not started until customer timing and API access are clear
PlannedHanwha WisenetFuture integration candidate
PlannedBosch / HoneywellTier-2 connectors planned
PlannedGeneric ONVIF eventsOptional enrichment alongside any source
LiveStays inside your network
Customer video does not leave the box. By default and by design.
CHEQIT runs on-prem next to the VMS. Connected services handle licensing, updates, and minimal health telemetry — never the camera footage. Air-gapped sites are activated with signed key files. Audit-grade decision context stays with each alert.
Plans
One on-prem product. Pick the plan that fits your estate.
Plans scale by camera capacity, activation mode, and operating model. Larger distributed fleets can move to custom terms without changing the installed product. Ordering opens by invitation during the limited release.
FAQ
Honest answers to the questions buyers ask first.
Does CHEQIT send my video to the cloud?
No. CHEQIT runs on-prem next to your VMS. Your video stays inside your network. The hosted services CHEQIT talks to handle licensing, software updates, and minimal health telemetry — not customer footage.
Can it run in an air-gapped environment?
Yes. Install CHEQIT, export a machine key from the local Licensing screen, and we issue a signed activation key file for that machine. Updates are delivered as signed bundles you transfer with your normal removable-media controls.
Which VMS systems are supported today?
Shinobi CE and direct ONVIF + RTSP are live in the MVP. Avigilon Unity discovery and API validation are in progress. Milestone XProtect, Genetec, Hanwha, Bosch, and Honeywell integrations are planned. See Connectors for current status, and tell us your VMS — we prioritise based on real customer deployments.
How does CHEQIT differ from “is the recorder online?” monitoring?
Most fleet monitors only check whether a device is reachable. CHEQIT checks whether the view is still useful — blur, freeze, darkness, scene drift, recording gaps, encoding faults — and attaches before/after evidence to every alert.
How long does deployment actually take?
An afternoon for the first site once you have a Linux host and credentials to the VMS. The signed bundle installs as a systemd service, you connect the VMS, and the first alerts come in as CHEQIT builds per-camera baselines.
What hardware do we need to host CHEQIT?
CHEQIT runs on a Linux x86_64 server or VM with systemd and SSD/NVMe storage. Headline tiers:
- Pilot / small site — 4 vCPU, 8 GB RAM, 100 GB SSD
- Standard site or leaf — 8 vCPU, 16 GB RAM, 250 GB SSD/NVMe
- Large site or federation primary — 16 vCPU, 32 GB RAM, 500 GB–1 TB NVMe
- Enterprise federated primary — 16–32 vCPU, 64 GB RAM, 1–2 TB NVMe
- HA standby — same shape as the active primary it stands in for
Treat these as starting points; full storage formulas, CPU/RAM/network guidance, and a slide-ready summary live in the System requirements & sizing doc.
Is there an alert-fatigue problem at fleet scale?
The pipeline is built to avoid it. Per-camera baselines, accepted orientation and scene-mode baselines, suppression windows, debounce, and camera/site alert policy keep noise controllable. Alerts are ranked into an operations queue, not flat email firehoses.
What does pricing look like for thousands of cameras?
Pro covers up to 512 cameras per instance; larger or federated estates use Enterprise terms. Self-serve checkout handles up to Pro; Enterprise is a short conversation. Annual billing typically saves around 17%.
Do you have SOC 2 / HIPAA / ISO 27001?
Not yet — we are honest about it. The current trust story is architectural: on-prem deployment, no video egress, signed releases, offline activation, scoped RBAC, and tamper-evident audit options. Formal certifications are on the roadmap once the product enters general availability.
What you get
Fewer surprises, one clear list.
See failures before they cost evidence
Surface blur, tampering, and signal loss while there's still time to fix them.
One ranked repair queue for security & IT
Both teams work the same prioritised list instead of trading blame.
Runs fully on-prem — no video egress
Monitoring lives inside your perimeter; footage stays on your network.
Offline / air-gapped activation
License and update with signed key files, even on isolated networks.