Fly-tipping
Detect arrivals, unusual stopping, after-hours activity, and dumped material at vulnerable locations.
SiteGuard Custom CCTV
Cameras that understand the requirement.
Combine mobile towers, permanent CCTV, resilient communications, and purpose-built analytics. The aim is not more footage. It is earlier insight and a clearer response.
CAMERAEDGE AIMULTI-NETWORK COMMS
01Mobile when the site moves
Solar towers, rapid deployment, wireless links, and relocatable coverage.
02Permanent when the risk stays
Surveyed CCTV, detection, recording, access, and long-term service.
03Custom when the question is specific
Models, rules, and reporting designed around the event that matters.
Custom analytics examples
Start with the behaviour, not the software.
Each system begins with a question: what should be noticed, how quickly, what evidence is useful, and who should receive it?
Traffic behaviour
Understand vehicle flow, queues, direction, dwell time, zone breaches, and near-miss patterns.
Insect density
Count visible insect activity and build time-based trends for agriculture, research, and ecology projects.
Wildlife identification
Classify likely species, record activity windows, and route uncertain events for human verification.
Wellbeing
Use privacy-aware occupancy, fall, isolation, and movement indicators to support agreed safeguarding workflows.
Baggage recognition
Flag left, removed, or unusually moved baggage and preserve a clear event trail for operator review.
Store actions
Measure queues, dwell, shelf interaction, and selected action patterns without turning the system into a generic camera feed.
The edge-AI pathway
Progressing towards NVIDIA Jetson development.
Jetson gives SiteGuard a practical route to run computer-vision models at the edge: close to the cameras, with lower bandwidth demand, faster event handling, and control over what leaves the site.
We will treat the development kit as a proving platform. Every production design still needs the right enclosure, power, storage, thermal management, cybersecurity, and support plan.
COMPUTE
Jetson-class edge board
Local inference, camera pipelines, model optimisation, and secure application containers.
COMMS
SIM & router strategy
Multi-network connectivity, managed data, VPN routing, health checks, and failover.
POWER
Designed for the field
Solar and battery budgets, controlled shutdown, remote restart, and thermal protection.
System architecture
From camera to action.
- 01 / Capture
Cameras & sensors
Fixed, PTZ, thermal, low-light, audio, environmental, or specialist inputs.
- 02 / Understand
Edge AI compute
Jetson-class processing runs the selected models close to the camera.
- 03 / Connect
Managed comms
Multi-network SIM, 4G/5G, Ethernet, Wi-Fi, VPN, and monitored fallback paths.
- 04 / Respond
Useful outputs
Verified alerts, evidence clips, counts, trends, integrations, and operator actions.
Built with controls
Useful analytics need responsible boundaries.
Every use case should define retention, privacy zones, human verification, performance testing, access controls, and the limits of the model before it becomes operational.
Plan a proof of concept ↗