CCTV & Surveillance

Edge AI runs analytics on the camera itself. Pros: lower network bandwidth, faster response time, works even if the server is offline. Cons: each camera is more expensive, and analytics features are limited to what the camera model supports. Server-side AI runs analytics on a central server processing streams from many cameras. Pros: more sophisticated cross-camera analytics like person re-identification across the site, easier to upgrade analytics independently, supports older cameras. Cons: higher bandwidth, server is a single point that needs redundancy. Most enterprise deployments use a hybrid — edge AI for time-critical alerts (intrusion, line crossing, ANPR) and server-side for investigation and complex multi-camera analysis.

Facial recognition is permitted under the Saudi PDPL when deployed for legitimate security purposes with explicit policy authorization, documented data handling procedures, and appropriate signage in monitored areas. It is most commonly deployed for access control (matching an authorized employee or visitor against an internal database) and watch-list monitoring in high-security environments. We deploy facial recognition only when a client has confirmed the legal basis under their operational scope and Saudi PDPL requirements, and we never deploy it for general public-space mass surveillance without specific regulatory authorization.

Several layers. (1) Camera network is VLAN-isolated from the corporate network. (2) Default passwords are changed and credentials are managed through a centralized authentication system. (3) Camera firmware is kept patched against known vulnerabilities, with a formal review schedule. (4) Camera traffic uses HTTPS/TLS where supported. (5) Management access requires multi-factor authentication. (6) System events are logged to the client's SIEM or to a dedicated audit log. (7) We follow manufacturer security baselines and Saudi NCA Essential Cybersecurity Controls (ECC) where applicable.

Retention is driven by client policy and regulatory requirement. Typical baselines: 30 days for general operational sites, 60 to 90 days for retail and corporate, 90 to 180 days for government and high-security, and longer (up to several years) for compliance-driven environments like banking. Storage sizing is calculated based on camera count, resolution, frame rate, codec efficiency (H.265 is the current standard, roughly half the storage of H.264), and motion-based recording where applicable. A typical 100-camera enterprise site at 4MP/15fps/H.265 with 60-day retention needs roughly 100 to 150 TB of storage. We model the calculation per project in the proposal.

Yes. We routinely integrate IP surveillance with major access control platforms (door events trigger camera tagging and clip retention), alarm systems (alarm activation pops the relevant camera on operator screens), building management (HVAC and lighting events correlated with camera feeds), and physical security information management (PSIM) platforms that consolidate everything into a single operator view. Integration scope is defined in the project agreement and tested against your live systems during commissioning.